docs/tutorials/federated_learning_with_differential_privacy.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1YRGLD1pOOrJ"
      },
      "source": [
        "##### Copyright 2021 The TensorFlow Federated Authors."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "cellView": "form",
        "id": "koW3R4ntOgLS"
      },
      "outputs": [],
      "source": [
        "#@title Licensed under the Apache License, Version 2.0 (the \"License\");\n",
        "# you may not use this file except in compliance with the License.\n",
        "# You may obtain a copy of the License at\n",
        "#\n",
        "# https://www.apache.org/licenses/LICENSE-2.0\n",
        "#\n",
        "# Unless required by applicable law or agreed to in writing, software\n",
        "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
        "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
        "# See the License for the specific language governing permissions and\n",
        "# limitations under the License."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "coAumH42q9nz"
      },
      "source": [
        "\u003ctable class=\"tfo-notebook-buttons\" align=\"left\"\u003e\n",
        "  \u003ctd\u003e\n",
        "    \u003ca target=\"_blank\" href=\"https://www.tensorflow.org/federated/tutorials/federated_learning_with_differential_privacy\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" /\u003eView on TensorFlow.org\u003c/a\u003e\n",
        "  \u003c/td\u003e\n",
        "  \u003ctd\u003e\n",
        "    \u003ca target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/federated/blob/v0.74.0/docs/tutorials/federated_learning_with_differential_privacy.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" /\u003eRun in Google Colab\u003c/a\u003e\n",
        "  \u003c/td\u003e\n",
        "  \u003ctd\u003e\n",
        "    \u003ca target=\"_blank\" href=\"https://github.com/tensorflow/federated/blob/v0.74.0/docs/tutorials/federated_learning_with_differential_privacy.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" /\u003eView source on GitHub\u003c/a\u003e\n",
        "  \u003c/td\u003e\n",
        "  \u003ctd\u003e\n",
        "    \u003ca href=\"https://storage.googleapis.com/tensorflow_docs/federated/docs/tutorials/federated_learning_with_differential_privacy.ipynb\"\u003e\u003cimg src=\"https://www.tensorflow.org/images/download_logo_32px.png\" /\u003eDownload notebook\u003c/a\u003e\n",
        "  \u003c/td\u003e\n",
        "\u003c/table\u003e"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "grBmytrShbUE"
      },
      "source": [
        "# Differential Privacy in TFF\n",
        "\n",
        "This tutorial will demonstrate the recommended best practice for training models with user-level Differential Privacy using Tensorflow Federated. We will use the DP-SGD algorithm of [Abadi et al., \"Deep Learning with Differential Privacy\"](https://arxiv.org/abs/1607.00133) modified for user-level DP in a federated context in [McMahan et al., \"Learning Differentially Private Recurrent Language Models\"](https://arxiv.org/abs/1710.06963).\n",
        "\n",
        "Differential Privacy (DP) is a widely used method for bounding and quantifying the privacy leakage of sensitive data when performing learning tasks. Training a model with user-level DP guarantees that the model is unlikely to learn anything significant about the data of any individual, but can still (hopefully!) learn patterns that exist in the data of many clients.\n",
        "\n",
        "We will train a model on the federated EMNIST dataset. There is an inherent trade-off between utility and privacy, and it may be difficult to train a model with high privacy that performs as well as a state-of-the-art non-private model. For expediency in this tutorial, we will train for just 100 rounds, sacrificing some quality in order to demonstrate how to train with high privacy. If we used more training rounds, we could certainly have a somewhat higher-accuracy private model, but not as high as a model trained without DP."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yiq_MY4LopET"
      },
      "source": [
        "## Before we begin\n",
        "\n",
        "First, let us make sure the notebook is connected to a backend that has the relevant components compiled. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ke7EyuvG0Zyn"
      },
      "outputs": [],
      "source": [
        "#@test {\"skip\": true}\n",
        "!pip install --quiet --upgrade dp-accounting\n",
        "!pip install --quiet --upgrade tensorflow-federated"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Cmp7VMf15aSu"
      },
      "source": [
        "Some imports we will need for the tutorial. We will use `tensorflow_federated`, the open-source framework for machine learning and other computations on decentralized data, as well as `dp_accounting`, an open-source library for analyzing differentially private algorithms."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "rtgStTrNIId-"
      },
      "outputs": [],
      "source": [
        "import collections\n",
        "\n",
        "import dp_accounting\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import tensorflow as tf\n",
        "import tensorflow_federated as tff"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X6eWsahmQpmi"
      },
      "source": [
        "Run the following \"Hello World\"\n",
        "example to make sure the TFF environment is correctly setup. If it doesn't work,\n",
        "please refer to the [Installation](../install.md) guide for instructions."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wjX3wmC-P1aE",
        "outputId": "b3e2cf36-a8f0-42af-84f0-24affeaad54a"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "b'Hello, World!'"
            ]
          },
          "execution_count": 4,
          "metadata": {
            "tags": []
          },
          "output_type": "execute_result"
        }
      ],
      "source": [
        "@tff.federated_computation\n",
        "def hello_world():\n",
        "  return 'Hello, World!'\n",
        "\n",
        "hello_world()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A8xl6I2X9ObS"
      },
      "source": [
        "## Download and preprocess the federated EMNIST dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "EdY72mGKJqi0"
      },
      "outputs": [],
      "source": [
        "def get_emnist_dataset():\n",
        "  emnist_train, emnist_test = tff.simulation.datasets.emnist.load_data(\n",
        "      only_digits=True)\n",
        "\n",
        "  def element_fn(element):\n",
        "    return collections.OrderedDict(\n",
        "        x=tf.expand_dims(element['pixels'], -1), y=element['label'])\n",
        "\n",
        "  def preprocess_train_dataset(dataset):\n",
        "    # Use buffer_size same as the maximum client dataset size,\n",
        "    # 418 for Federated EMNIST\n",
        "    return (dataset.map(element_fn)\n",
        "                   .shuffle(buffer_size=418)\n",
        "                   .repeat(1)\n",
        "                   .batch(32, drop_remainder=False))\n",
        "\n",
        "  def preprocess_test_dataset(dataset):\n",
        "    return dataset.map(element_fn).batch(128, drop_remainder=False)\n",
        "\n",
        "  emnist_train = emnist_train.preprocess(preprocess_train_dataset)\n",
        "  emnist_test = preprocess_test_dataset(\n",
        "      emnist_test.create_tf_dataset_from_all_clients())\n",
        "  return emnist_train, emnist_test\n",
        "\n",
        "train_data, test_data = get_emnist_dataset()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3ntJ5coIJxS2"
      },
      "source": [
        "## Define our model."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YK_UGq_0KGMX"
      },
      "outputs": [],
      "source": [
        "def my_model_fn():\n",
        "  model = tf.keras.models.Sequential([\n",
        "      tf.keras.layers.Reshape(input_shape=(28, 28, 1), target_shape=(28 * 28,)),\n",
        "      tf.keras.layers.Dense(200, activation=tf.nn.relu),\n",
        "      tf.keras.layers.Dense(200, activation=tf.nn.relu),\n",
        "      tf.keras.layers.Dense(10)])\n",
        "  return tff.learning.models.from_keras_model(\n",
        "      keras_model=model,\n",
        "      loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
        "      input_spec=test_data.element_spec,\n",
        "      metrics=[tf.keras.metrics.SparseCategoricalAccuracy()])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-2jnGd-CKTYW"
      },
      "source": [
        "## Determine the noise sensitivity of the model.\n",
        "\n",
        "To get user-level DP guarantees, we must change the basic Federated Averaging algorithm in two ways. First, the clients' model updates must be clipped before transmission to the server, bounding the maximum influence of any one client. Second, the server must add enough noise to the sum of user updates before averaging to obscure the worst-case client influence.\n",
        "\n",
        "For clipping, we use the adaptive clipping method of [Andrew et al. 2021, Differentially Private Learning with Adaptive Clipping](https://arxiv.org/abs/1905.03871), so no clipping norm needs to be explicitly set.\n",
        "\n",
        "Adding noise will in general degrade the utility of the model, but we can control the amount of noise in the average update at each round with two knobs: the standard deviation of the Gaussian noise added to the sum, and the number of clients in the average. Our strategy will be to first determine how much noise the model can tolerate with a relatively small number of clients per round with acceptable loss to model utility. Then to train the final model, we can increase the amount of noise in the sum, while proportionally scaling up the number of clients per round (assuming the dataset is large enough to support that many clients per round). This is unlikely to significantly affect model quality, since the only effect is to decrease variance due to client sampling (indeed we will verify that it does not in our case).\n",
        "\n",
        "To that end, we first train a series of models with 50 clients per round, with increasing amounts of noise. Specifically, we increase the \"noise_multiplier\" which is the ratio of the noise standard deviation to the clipping norm. Since we are using adaptive clipping, this means that the actual magnitude of the noise changes from round to round."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6-BZ_L4GMmXP"
      },
      "outputs": [],
      "source": [
        "total_clients = len(train_data.client_ids)\n",
        "\n",
        "def train(rounds, noise_multiplier, clients_per_round, data_frame):\n",
        "  # Using the `dp_aggregator` here turns on differential privacy with adaptive\n",
        "  # clipping.\n",
        "  aggregation_factory = tff.learning.model_update_aggregator.dp_aggregator(\n",
        "      noise_multiplier, clients_per_round)\n",
        "\n",
        "  # We use Poisson subsampling which gives slightly tighter privacy guarantees\n",
        "  # compared to having a fixed number of clients per round. The actual number of\n",
        "  # clients per round is stochastic with mean clients_per_round.\n",
        "  sampling_prob = clients_per_round / total_clients\n",
        "\n",
        "  # Build a federated averaging process.\n",
        "  # Typically a non-adaptive server optimizer is used because the noise in the\n",
        "  # updates can cause the second moment accumulators to become very large\n",
        "  # prematurely.\n",
        "  learning_process = tff.learning.algorithms.build_unweighted_fed_avg(\n",
        "        my_model_fn,\n",
        "        client_optimizer_fn=lambda: tf.keras.optimizers.SGD(0.01),\n",
        "        server_optimizer_fn=lambda: tf.keras.optimizers.SGD(1.0, momentum=0.9),\n",
        "        model_aggregator=aggregation_factory)\n",
        "\n",
        "  eval_process = tff.learning.build_federated_evaluation(my_model_fn)\n",
        "    \n",
        "  # Training loop.\n",
        "  state = learning_process.initialize()\n",
        "  for round in range(rounds):\n",
        "    if round % 5 == 0:\n",
        "      model_weights = learning_process.get_model_weights(state)\n",
        "      metrics = eval_process(model_weights, [test_data])['eval']\n",
        "      if round \u003c 25 or round % 25 == 0:\n",
        "        print(f'Round {round:3d}: {metrics}')\n",
        "      data_frame = data_frame.append({'Round': round,\n",
        "                                      'NoiseMultiplier': noise_multiplier,\n",
        "                                      **metrics}, ignore_index=True)\n",
        "\n",
        "    # Sample clients for a round. Note that if your dataset is large and\n",
        "    # sampling_prob is small, it would be faster to use gap sampling.\n",
        "    x = np.random.uniform(size=total_clients)\n",
        "    sampled_clients = [\n",
        "        train_data.client_ids[i] for i in range(total_clients)\n",
        "        if x[i] \u003c sampling_prob]\n",
        "    sampled_train_data = [\n",
        "        train_data.create_tf_dataset_for_client(client)\n",
        "        for client in sampled_clients]\n",
        "\n",
        "    # Use selected clients for update.\n",
        "    result = learning_process.next(state, sampled_train_data)\n",
        "    state = result.state\n",
        "    metrics = result.metrics\n",
        "\n",
        "  model_weights = learning_process.get_model_weights(state)\n",
        "  metrics = eval_process(model_weights, [test_data])['eval']\n",
        "  print(f'Round {rounds:3d}: {metrics}')\n",
        "  data_frame = data_frame.append({'Round': rounds,\n",
        "                                  'NoiseMultiplier': noise_multiplier,\n",
        "                                  **metrics}, ignore_index=True)\n",
        "\n",
        "  return data_frame\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qcaBxl0AbLTQ",
        "outputId": "22cdad6b-3ce2-4db0-b0be-ba1fe9599f8c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Starting training with noise multiplier: 0.0\n",
            "Round   0: OrderedDict([('sparse_categorical_accuracy', 0.112289384), ('loss', 2.5190482)])\n",
            "Round   5: OrderedDict([('sparse_categorical_accuracy', 0.19075724), ('loss', 2.2449977)])\n",
            "Round  10: OrderedDict([('sparse_categorical_accuracy', 0.18115693), ('loss', 2.163907)])\n",
            "Round  15: OrderedDict([('sparse_categorical_accuracy', 0.49970612), ('loss', 2.01017)])\n",
            "Round  20: OrderedDict([('sparse_categorical_accuracy', 0.5333317), ('loss', 1.8350543)])\n",
            "Round  25: OrderedDict([('sparse_categorical_accuracy', 0.5828517), ('loss', 1.6551636)])\n",
            "Round  50: OrderedDict([('sparse_categorical_accuracy', 0.7352077), ('loss', 0.8700141)])\n",
            "Round  75: OrderedDict([('sparse_categorical_accuracy', 0.7769152), ('loss', 0.6992781)])\n",
            "Round 100: OrderedDict([('sparse_categorical_accuracy', 0.8049814), ('loss', 0.62453026)])\n",
            "\n",
            "Starting training with noise multiplier: 0.5\n",
            "Round   0: OrderedDict([('sparse_categorical_accuracy', 0.09526841), ('loss', 2.4332638)])\n",
            "Round   5: OrderedDict([('sparse_categorical_accuracy', 0.20128821), ('loss', 2.2664592)])\n",
            "Round  10: OrderedDict([('sparse_categorical_accuracy', 0.35472178), ('loss', 2.130336)])\n",
            "Round  15: OrderedDict([('sparse_categorical_accuracy', 0.5480995), ('loss', 1.9713942)])\n",
            "Round  20: OrderedDict([('sparse_categorical_accuracy', 0.42246276), ('loss', 1.8045483)])\n",
            "Round  25: OrderedDict([('sparse_categorical_accuracy', 0.624902), ('loss', 1.4785467)])\n",
            "Round  50: OrderedDict([('sparse_categorical_accuracy', 0.7265625), ('loss', 0.85801566)])\n",
            "Round  75: OrderedDict([('sparse_categorical_accuracy', 0.77720904), ('loss', 0.70615387)])\n",
            "Round 100: OrderedDict([('sparse_categorical_accuracy', 0.7702537), ('loss', 0.72331005)])\n",
            "\n",
            "Starting training with noise multiplier: 0.75\n",
            "Round   0: OrderedDict([('sparse_categorical_accuracy', 0.098672606), ('loss', 2.422002)])\n",
            "Round   5: OrderedDict([('sparse_categorical_accuracy', 0.11794671), ('loss', 2.2227976)])\n",
            "Round  10: OrderedDict([('sparse_categorical_accuracy', 0.3208513), ('loss', 2.083766)])\n",
            "Round  15: OrderedDict([('sparse_categorical_accuracy', 0.49752644), ('loss', 1.8728142)])\n",
            "Round  20: OrderedDict([('sparse_categorical_accuracy', 0.5816761), ('loss', 1.6084186)])\n",
            "Round  25: OrderedDict([('sparse_categorical_accuracy', 0.62896746), ('loss', 1.378527)])\n",
            "Round  50: OrderedDict([('sparse_categorical_accuracy', 0.73153406), ('loss', 0.8705139)])\n",
            "Round  75: OrderedDict([('sparse_categorical_accuracy', 0.7789724), ('loss', 0.7113147)])\n",
            "Round 100: OrderedDict([('sparse_categorical_accuracy', 0.70944357), ('loss', 0.89495045)])\n",
            "\n",
            "Starting training with noise multiplier: 1.0\n",
            "Round   0: OrderedDict([('sparse_categorical_accuracy', 0.12002841), ('loss', 2.60482)])\n",
            "Round   5: OrderedDict([('sparse_categorical_accuracy', 0.104574844), ('loss', 2.3388205)])\n",
            "Round  10: OrderedDict([('sparse_categorical_accuracy', 0.29966694), ('loss', 2.089262)])\n",
            "Round  15: OrderedDict([('sparse_categorical_accuracy', 0.4067398), ('loss', 1.9109797)])\n",
            "Round  20: OrderedDict([('sparse_categorical_accuracy', 0.5123677), ('loss', 1.6472703)])\n",
            "Round  25: OrderedDict([('sparse_categorical_accuracy', 0.56416535), ('loss', 1.4362282)])\n",
            "Round  50: OrderedDict([('sparse_categorical_accuracy', 0.62323666), ('loss', 1.1682972)])\n",
            "Round  75: OrderedDict([('sparse_categorical_accuracy', 0.55968356), ('loss', 1.4779186)])\n",
            "Round 100: OrderedDict([('sparse_categorical_accuracy', 0.382837), ('loss', 1.9680436)])\n",
            "\n"
          ]
        }
      ],
      "source": [
        "data_frame = pd.DataFrame()\n",
        "rounds = 100\n",
        "clients_per_round = 50\n",
        "\n",
        "for noise_multiplier in [0.0, 0.5, 0.75, 1.0]:\n",
        "  print(f'Starting training with noise multiplier: {noise_multiplier}')\n",
        "  data_frame = train(rounds, noise_multiplier, clients_per_round, data_frame)\n",
        "  print()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x8h5cZ2OUmUF"
      },
      "source": [
        "Now we can visualize the evaluation set accuracy and loss of those runs."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "EHKzgJiQSxAE"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "def make_plot(data_frame):\n",
        "  plt.figure(figsize=(15, 5))\n",
        "\n",
        "  dff = data_frame.rename(\n",
        "      columns={'sparse_categorical_accuracy': 'Accuracy', 'loss': 'Loss'})\n",
        "\n",
        "  plt.subplot(121)\n",
        "  sns.lineplot(data=dff, x='Round', y='Accuracy', hue='NoiseMultiplier', palette='dark')\n",
        "  plt.subplot(122)\n",
        "  sns.lineplot(data=dff, x='Round', y='Loss', hue='NoiseMultiplier', palette='dark')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xiF8a2XxMt_8",
        "outputId": "dee18ef1-abca-40df-8bce-2d8eb6959d32"
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAE9CAYAAABZZMC4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90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\nOvF/E/rPzWgbLkQAAAAASUVORK5CYII=\n",
            "text/plain": [
              "\u003cFigure size 1500x500 with 2 Axes\u003e"
            ]
          },
          "metadata": {
            "tags": []
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "make_plot(data_frame)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3jeTMaZ5MunO"
      },
      "source": [
        "It appears that with 50 expected clients per round, this model can tolerate a noise multiplier of up to 0.5 without degrading model quality. A noise multiplier of 0.75 seems to cause a bit of model degradation, and 1.0 makes the model diverge.\n",
        "\n",
        "There is typically a tradeoff between model quality and privacy. The higher noise we use, the more privacy we can get for the same amount of training time and number of clients. Conversely, with less noise, we may have a more accurate model, but we'll have to train with more clients per round to reach our target privacy level.\n",
        "\n",
        "With the experiment above, we might decide that the small amount of model deterioration at 0.75 is acceptable in order to train the final model faster, but let's assume we want to match the performance of the 0.5 noise-multiplier model.\n",
        "\n",
        "Now we can use dp_accounting functions to determine how many expected clients per round we would need to get acceptable privacy. Standard practice is to choose delta somewhat smaller than one over the number of records in the dataset. This dataset has 3383 total training users, so let's aim for (2, 1e-5)-DP.\n",
        "\n",
        "We use `dp_accounting.calibrate_dp_mechanism` to search over the number of clients per round. The privacy accountant (`RdpAccountant`) we use to estimate privacy given a `dp_accounting.DpEvent` is based on [Wang et al. (2018)](https://arxiv.org/abs/1808.00087) and [Mironov et al. (2019)](https://arxiv.org/pdf/1908.10530.pdf)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "executionInfo": {
          "elapsed": 3437,
          "status": "ok",
          "timestamp": 1671496851266,
          "user": {
            "displayName": "",
            "userId": ""
          },
          "user_tz": 480
        },
        "id": "QDMUmKWKXAMB",
        "outputId": "d32d504f-2108-41c7-9b4e-ac07e238985d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "To get (2, 1e-05)-DP, use 120 clients with noise multiplier 1.2.\n"
          ]
        }
      ],
      "source": [
        "total_clients = 3383\n",
        "noise_to_clients_ratio = 0.01\n",
        "target_delta = 1e-5\n",
        "target_eps = 2\n",
        "\n",
        "# Initialize arguments to dp_accounting.calibrate_dp_mechanism.\n",
        "\n",
        "# No-arg callable that returns a fresh accountant.\n",
        "make_fresh_accountant = dp_accounting.rdp.RdpAccountant\n",
        "\n",
        "# Create function that takes expected clients per round and returns a \n",
        "# dp_accounting.DpEvent representing the full training process.\n",
        "def make_event_from_param(clients_per_round):\n",
        "  q = clients_per_round / total_clients\n",
        "  noise_multiplier = clients_per_round * noise_to_clients_ratio\n",
        "  gaussian_event = dp_accounting.GaussianDpEvent(noise_multiplier)\n",
        "  sampled_event = dp_accounting.PoissonSampledDpEvent(q, gaussian_event)\n",
        "  composed_event = dp_accounting.SelfComposedDpEvent(sampled_event, rounds)\n",
        "  return composed_event\n",
        "\n",
        "# Create object representing the search range [1, 3383].\n",
        "bracket_interval = dp_accounting.ExplicitBracketInterval(1, total_clients)\n",
        "\n",
        "# Perform search for smallest clients_per_round achieving the target privacy.\n",
        "clients_per_round = dp_accounting.calibrate_dp_mechanism(\n",
        "    make_fresh_accountant, make_event_from_param, target_eps, target_delta,\n",
        "    bracket_interval, discrete=True\n",
        ")\n",
        "\n",
        "noise_multiplier = clients_per_round * noise_to_clients_ratio\n",
        "print(f'To get ({target_eps}, {target_delta})-DP, use {clients_per_round} '\n",
        "      f'clients with noise multiplier {noise_multiplier}.')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ExhsP4nxiaok"
      },
      "source": [
        "Now we can train our final private model for release.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "aIacNzuxibOB",
        "outputId": "28fa2cc5-b760-4856-d9bc-1bbd8450f5fd"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Round   0: OrderedDict([('sparse_categorical_accuracy', 0.08260678), ('loss', 2.6334999)])\n",
            "Round   5: OrderedDict([('sparse_categorical_accuracy', 0.1492212), ('loss', 2.259542)])\n",
            "Round  10: OrderedDict([('sparse_categorical_accuracy', 0.28847474), ('loss', 2.155699)])\n",
            "Round  15: OrderedDict([('sparse_categorical_accuracy', 0.3989518), ('loss', 2.0156953)])\n",
            "Round  20: OrderedDict([('sparse_categorical_accuracy', 0.5086697), ('loss', 1.8261365)])\n",
            "Round  25: OrderedDict([('sparse_categorical_accuracy', 0.6204692), ('loss', 1.5602393)])\n",
            "Round  50: OrderedDict([('sparse_categorical_accuracy', 0.70008814), ('loss', 0.91155165)])\n",
            "Round  75: OrderedDict([('sparse_categorical_accuracy', 0.78421336), ('loss', 0.6820159)])\n",
            "Round 100: OrderedDict([('sparse_categorical_accuracy', 0.7955525), ('loss', 0.6585961)])\n"
          ]
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAE9CAYAAABZZMC4AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90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\nTfMlgwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "\u003cFigure size 1500x500 with 2 Axes\u003e"
            ]
          },
          "metadata": {
            "tags": []
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "rounds = 100\n",
        "noise_multiplier = 1.2\n",
        "clients_per_round = 120\n",
        "\n",
        "data_frame = pd.DataFrame()\n",
        "data_frame = train(rounds, noise_multiplier, clients_per_round, data_frame)\n",
        "\n",
        "make_plot(data_frame)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w-N7cvpVkFR6"
      },
      "source": [
        "As we can see, the final model has similar loss and accuracy to the model trained without noise, but this one satisfies (2, 1e-5)-DP."
      ]
    }
  ],
  "metadata": {
    "colab": {
      "name": "federated_learning_with_differential_privacy.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}