tf_placeholder.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "tf_placeholder.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/dauparas/tensorflow_examples/blob/master/tf_placeholder.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "H-zFSs0FbbSL",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#Import libraries\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "plt.style.use('ggplot')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "IvTijQ2Td9rn",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#Create some data\n",
        "np.random.seed(0)\n",
        "x = np.random.uniform(0,1,100) \n",
        "y = 1.2*x + 0.1*np.random.randn(x.shape[0])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "NMiIYqrveU_1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "aa0b5954-02ae-4d1f-e421-47342fa8f732"
      },
      "cell_type": "code",
      "source": [
        "#Plot the data\n",
        "plt.scatter(x, y);"
      ],
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD4CAYAAADxeG0DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "1wVUk8XreZpk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "#Reset the graph \n",
        "tf.reset_default_graph()\n",
        "\n",
        "#Parameters\n",
        "learning_rate = 0.01\n",
        "batch_size = 10\n",
        "n_epochs = 100\n",
        "\n",
        "#Create placeholders\n",
        "X = tf.placeholder(tf.float32, shape=[None, 1], name='X')\n",
        "Y = tf.placeholder(tf.float32, shape=[None, 1], name='Y')\n",
        "LEARNING_RATE = tf.placeholder(tf.float32, shape=[], name='LEARNING_RATE')\n",
        "#Create variables\n",
        "A = tf.get_variable('A', shape=[1], dtype=tf.float32, initializer=tf.random_normal_initializer())\n",
        "B = tf.get_variable('B', shape=[1], dtype=tf.float32, initializer=tf.zeros_initializer())\n",
        "\n",
        "#Create a linear model\n",
        "y_pred = A*X + B\n",
        "\n",
        "#Define a loss function\n",
        "loss = tf.reduce_sum(tf.square(y_pred-Y), axis=0)\n",
        "\n",
        "#Define an optimizer\n",
        "optimizer = tf.train.AdamOptimizer(LEARNING_RATE).minimize(loss)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "tqbP-efY8sdU",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def batch_generator(features, labels, batch_size):\n",
        "  \"\"\"Function to create python generator to shuffle and split features and labels into batches along the first dimension.\"\"\"\n",
        "  idx = np.arange(features.shape[0])\n",
        "  np.random.shuffle(idx)\n",
        "  for start_idx in range(0, features.shape[0], batch_size):\n",
        "    end_idx = min(start_idx + batch_size, features.shape[0])\n",
        "    part = idx[start_idx:end_idx]\n",
        "    yield features[part], labels[part]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "9zxjy9n6A577",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        },
        "outputId": "68f48dc6-4556-4854-e06b-0d941992616e"
      },
      "cell_type": "code",
      "source": [
        "#Create a TF session:\n",
        "saver = tf.train.Saver()\n",
        "with tf.Session() as sess:\n",
        "  sess.run(tf.global_variables_initializer())\n",
        "  for i in range(n_epochs):\n",
        "    learning_rate = 0.1 - 0.05*((i+1)/n_epochs) #variable learning rate\n",
        "    batch_size = np.int(2+20*((i+1)/n_epochs)) #variable batch size\n",
        "    gen = batch_generator(x, y, batch_size) #create batch generator\n",
        "    total_loss = 0\n",
        "    for j in range(np.int(x.shape[0]/batch_size)):\n",
        "      x_batch, y_batch = gen.__next__()\n",
        "      x_batch, y_batch = x_batch[:, np.newaxis], y_batch[:, np.newaxis]\n",
        "      _, batch_loss = sess.run([optimizer, loss], feed_dict={X: x_batch, Y: y_batch, LEARNING_RATE: learning_rate})\n",
        "      total_loss += batch_loss\n",
        "    if (i+1)% 10 == 0 or i == 0:\n",
        "      print('Epoch: {0}, Loss: {1}'.format((i+1), total_loss))\n",
        "    a, b = sess.run([A, B])\n",
        "  save_path = saver.save(sess, \"./tmp/linear.ckpt\")\n",
        "  print(\"Model saved in path: %s\" % save_path)"
      ],
      "execution_count": 159,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch: 1, Loss: [3.9511735]\n",
            "Epoch: 10, Loss: [1.1477625]\n",
            "Epoch: 20, Loss: [1.0226266]\n",
            "Epoch: 30, Loss: [1.1828575]\n",
            "Epoch: 40, Loss: [1.1779041]\n",
            "Epoch: 50, Loss: [1.2187423]\n",
            "Epoch: 60, Loss: [1.1227932]\n",
            "Epoch: 70, Loss: [0.990875]\n",
            "Epoch: 80, Loss: [0.9205338]\n",
            "Epoch: 90, Loss: [1.1026571]\n",
            "Epoch: 100, Loss: [0.89151335]\n",
            "Model saved in path: ./tmp/linear.ckpt\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "AoJHhvm_F9-w",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "outputId": "afd31cf7-3f7b-4d42-bb9f-28894c440bc0"
      },
      "cell_type": "code",
      "source": [
        "plt.scatter(x, y);\n",
        "plt.scatter(x, a*x+b);\n",
        "print(a, b)"
      ],
      "execution_count": 164,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[1.1886508] [0.03255336]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD4CAYAAADxeG0DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJJALIUSO+/8BlIW7Eoo0hkQAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "JTvuUwCpHSJU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "2cae08ca-5292-402c-8c7e-4a06c5c81a6b"
      },
      "cell_type": "code",
      "source": [
        "tf.reset_default_graph()\n",
        "\n",
        "# Create some variables.\n",
        "A = tf.get_variable('A', shape=[1], dtype=tf.float32)\n",
        "B = tf.get_variable('B', shape=[1], dtype=tf.float32)\n",
        "\n",
        "# Add ops to save and restore all the variables.\n",
        "saver = tf.train.Saver()\n",
        "\n",
        "# Later, launch the model, use the saver to restore variables from disk, and\n",
        "# do some work with the model.\n",
        "with tf.Session() as sess:\n",
        "  # Restore variables from disk.\n",
        "  saver.restore(sess, \"./tmp/linear.ckpt\")\n",
        "  print(\"Model restored.\")\n",
        "  # Check the values of the variables\n",
        "  print(\"A : %s\" % A.eval())\n",
        "  print(\"B : %s\" % B.eval())"
      ],
      "execution_count": 165,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Restoring parameters from ./tmp/linear.ckpt\n",
            "Model restored.\n",
            "A : [1.1886508]\n",
            "B : [0.03255336]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "oC_pi0bwLETM",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}