RBM (1).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "RBM.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "LmUTKu6g8GY5",
        "colab_type": "code",
        "outputId": "272de7d2-389e-4c38-805b-e34f0d7c05e8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 330
        }
      },
      "source": [
        "import tensorflow as tf\n",
        "import os \n",
        "import numpy as np\n",
        "from tensorflow.examples.tutorials.mnist import input_data\n",
        "from tqdm import tqdm\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from rbm import RBM, sample"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ModuleNotFoundError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-7-f686eadde8e0>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpyplot\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mrbm\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mRBM\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msample\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'rbm'",
            "",
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0;32m\nNOTE: If your import is failing due to a missing package, you can\nmanually install dependencies using either !pip or !apt.\n\nTo view examples of installing some common dependencies, click the\n\"Open Examples\" button below.\n\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v6WA6c-k8J4-",
        "colab_type": "code",
        "outputId": "fe015100-24e3-44cf-9d5b-47add9d336e3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 218
        }
      },
      "source": [
        "pip install \"git+https://github.com/chimera0/accel-brain-code.git#egg=pydbm&subdirectory=Deep-Learning-by-means-of-Design-Pattern\""
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting pydbm\n",
            "  Cloning https://github.com/chimera0/accel-brain-code.git to /tmp/pip-install-3d33ayfj/pydbm\n",
            "  Running command git clone -q https://github.com/chimera0/accel-brain-code.git /tmp/pip-install-3d33ayfj/pydbm\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from pydbm) (1.18.2)\n",
            "Requirement already satisfied: cython in /usr/local/lib/python3.6/dist-packages (from pydbm) (0.29.15)\n",
            "Building wheels for collected packages: pydbm\n",
            "  Building wheel for pydbm (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pydbm: filename=pydbm-1.5.2-cp36-cp36m-linux_x86_64.whl size=17505006 sha256=fb2d1f67c32f1d4b919f852a4d2ee891baea2df1cbc3bbbf530fb4d4bb12096e\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-tv4sii8a/wheels/b1/dd/56/71e6f0365c3e24d4c192e4c824334cc475486fc60b93a8c772\n",
            "Successfully built pydbm\n",
            "Installing collected packages: pydbm\n",
            "Successfully installed pydbm-1.5.2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RwxPhHCO8wPz",
        "colab_type": "code",
        "outputId": "e1c11843-f610-4d5a-adf8-6712a3a9a7be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        }
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tfrbm import BBRBM, GBRBM\n",
        "from tensorflow.examples.tutorials.mnist import input_data\n",
        "\n",
        "mnist = input_data.read_data_sets('MNIST_data/', one_hot=True)\n",
        "mnist_images = mnist.train.images\n",
        "\n",
        "bbrbm = BBRBM(n_visible=784, n_hidden=64, learning_rate=0.01, momentum=0.95, use_tqdm=True)\n",
        "errs = bbrbm.fit(mnist_images, n_epoches=30, batch_size=10)\n",
        "plt.plot(errs)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ModuleNotFoundError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-5-07d531256a41>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnumpy\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpyplot\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mtfrbm\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mBBRBM\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mGBRBM\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexamples\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtutorials\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmnist\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0minput_data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'tfrbm'",
            "",
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0;32m\nNOTE: If your import is failing due to a missing package, you can\nmanually install dependencies using either !pip or !apt.\n\nTo view examples of installing some common dependencies, click the\n\"Open Examples\" button below.\n\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AE4w6o5-_KNc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "def sample(probabilities, mode='bernoulli'):\n",
        "    ''' Sample a tensor based on the probabilities (A tensor given by get_probabilities)'''\n",
        "    if mode=='bernoulli':\n",
        "        return tf.floor(probabilities + tf.random_uniform(tf.shape(probabilities), 0, 1))\n",
        "    elif mode=='gaussian':\n",
        "        return tf.add(probabilities, tf.random_normal(tf.shape(probabilities), mean=0.0, stddev=1.)) # Add noise to the original probabilities\n",
        "\n",
        "\n",
        "\n",
        "# Simple RBM class\n",
        "class RBM:\n",
        "\n",
        "    def __init__(self, n_visible, n_hidden, lr, epochs, mode='bernoulli'):\n",
        "        ''' Initialize a model for an RBM with one layer of hidden units '''\n",
        "        self.mode = mode # bernoulli or gaussian RBM\n",
        "        self.n_hidden = n_hidden #  Number of hidden nodes\n",
        "        self.n_visible = n_visible # Number of visible nodes\n",
        "        self.lr = lr # Learning rate for the CD algorithm\n",
        "        self.epochs = epochs # Number of iterations to run the algorithm for\n",
        "\n",
        "        # Initialize weights and biases\n",
        "        with tf.name_scope('Weights'):\n",
        "            self.W = tf.Variable(tf.random_normal([self.n_visible, self.n_hidden], mean=0., stddev=4 * np.sqrt(6. / (self.n_visible + self.n_hidden))), name=\"weights\")\n",
        "        tf.summary.histogram('weights',self.W)\n",
        "        self.vb = tf.Variable(tf.zeros([1, self.n_visible]),tf.float32, name=\"visible_bias\")\n",
        "        self.hb = tf.Variable(tf.zeros([1, self.n_hidden]),tf.float32, name=\"hidden_bias\")\n",
        "\n",
        "\n",
        "    def get_probabilities(self, layer, val):\n",
        "        ''' Return a tensor of probabilities associated with the layer specified'''\n",
        "        if layer == 'hidden':\n",
        "            with tf.name_scope(\"Hidden_Probabilities\"):\n",
        "                if self.mode=='bernoulli':\n",
        "                    return tf.nn.sigmoid(tf.matmul(val, self.W) + self.hb)\n",
        "                elif self.mode=='gaussian':\n",
        "                    return tf.matmul(val, self.W) + self.hb\n",
        "\n",
        "        elif layer == 'visible':\n",
        "            with tf.name_scope(\"Visible_Probabilities\"):\n",
        "                return tf.nn.sigmoid(tf.matmul(val, tf.transpose(self.W)) + self.vb)\n",
        "\n",
        "\n",
        "    def CD(self, v, K=1):\n",
        "        ''' K-step Contrastive Divergence using Gibbs sampling. Return parameters update. '''\n",
        "        with tf.name_scope(\"Contrastive_Divergence\"):\n",
        "            h_prob = self.get_probabilities('hidden', v)\n",
        "            h_state = sample(h_prob, mode=self.mode)\n",
        "            pos_divergence = tf.matmul(tf.transpose(v), h_prob) # Positive Divergence + h(v).v^T\n",
        "\n",
        "            fake_v_prob = self.get_probabilities('visible', h_state)\n",
        "            fake_v_state = fake_v_prob #sample(fake_v_prob)\n",
        "\n",
        "            fake_h_prob = self.get_probabilities('hidden', fake_v_state)\n",
        "            fake_h_state = sample(fake_h_prob, mode=self.mode)\n",
        "\n",
        "            for i in range(K-1): # Number of steps to run the algorithm\n",
        "\n",
        "                fake_v_prob = self.get_probabilities('visible', fake_h_state)\n",
        "                fake_v_state = fake_v_prob #sample(fake_v_prob)\n",
        "\n",
        "                fake_h_prob = self.get_probabilities('hidden', fake_v_state)\n",
        "                fake_h_state = sample(fake_h_prob, mode=self.mode)\n",
        "\n",
        "            neg_divergence = tf.matmul(tf.transpose(fake_v_state), fake_h_prob) # Negative Divergence - h(v').v'^T\n",
        "\n",
        "            dW = pos_divergence-neg_divergence\n",
        "            dvb = v-fake_v_state\n",
        "            dhb = h_prob-fake_h_prob\n",
        "\n",
        "            # Similarity between reconstructed visible layer and input during training. \n",
        "            self.rec_error = tf.reduce_mean(tf.squared_difference(v, fake_v_state))\n",
        "            tf.summary.scalar('reconstruction_error', self.rec_error)\n",
        "\n",
        "            self.div = tf.reduce_mean(tf.abs(dW))\n",
        "            tf.summary.scalar('weights_increment', self.div)\n",
        "\n",
        "            return dW, dvb, dhb\n",
        "\n",
        "\n",
        "    def update(self, v, K=1):\n",
        "        batch_size = tf.cast(tf.shape(v)[0], tf.float32) # batch size\n",
        "        dW, dvb, dhb = self.CD(v, K=K) # contrastive divergence\n",
        "\n",
        "        delta_w = (self.lr/batch_size)*dW # weight gradient\n",
        "        delta_vb = (self.lr/batch_size)*(tf.reduce_sum(dvb, 0, keep_dims=True)) # visible bias gradient\n",
        "        delta_hb = (self.lr/batch_size)*(tf.reduce_sum(dhb, 0, keep_dims=True)) # hidden bias gradient\n",
        "\n",
        "        train_op = [self.W.assign_add(delta_w), self.vb.assign_add(delta_vb), self.hb.assign_add(delta_hb)] \n",
        "        return train_op\n",
        "\n",
        "\n",
        "    def gibbs(self, steps, v):\n",
        "        ''' Use the Gibbs sampler for a network of hidden and visible units. Return a sampled version of the input'''\n",
        "        with tf.name_scope(\"Gibbs_sampling\"):\n",
        "            for i in range(steps): # Number of steps to run the algorithm\n",
        "                hidden_p = self.get_probabilities('hidden', v) # v: input data\n",
        "                h = sample(hidden_p, mode=self.mode)\n",
        "\n",
        "                visible_p = self.get_probabilities('visible', h)\n",
        "                v = visible_p\n",
        "                #v = sample(visible_p)\n",
        "            return visible_p\n",
        "\n",
        "\n",
        "    def free_energy(self, v):\n",
        "        ''' Compute the free energy for a visible state'''\n",
        "        vbias_term = tf.matmul(v, tf.transpose(self.vb))\n",
        "        x_b = tf.matmul(v, self.W) + self.hb\n",
        "        hidden_term = tf.reduce_sum(tf.log(1 + tf.exp(x_b)))\n",
        "        return - hidden_term - vbias_term\n",
        "\n",
        "\n",
        "    def get_feature_map(self):\n",
        "        ''' Return hidden features'''\n",
        "        ft_map = {}\n",
        "        for k in range(self.n_hidden):\n",
        "            ft_map[k] = self.get_probabilities('visible', tf.expand_dims(tf.one_hot(k+1, self.n_hidden),0))\n",
        "        return ft_map"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DjtgzvCWGM9c",
        "colab_type": "code",
        "outputId": "bea69f25-1f41-411e-9b41-c4b1adb6a753",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 524
        }
      },
      "source": [
        "mnist = input_data.read_data_sets('MNIST_data', one_hot = True)\n",
        "\n",
        "# Initialize RBM Model for MNIST reconstruction\n",
        "rbm_model = RBM(n_visible = 784, n_hidden = 144, lr = tf.constant(0.05, tf.float32), epochs = 10000, mode='bernoulli')\n",
        "#rbm_model = RBM(n_visible = 784, n_hidden = 144, lr = tf.constant(0.003, tf.float32), epochs = 10000, mode='gaussian')\n",
        "\n",
        "# Placeholder for the visible layer of the RBM computation graph.\n",
        "v = tf.placeholder(tf.float32, shape=[None, rbm_model.n_visible], name=\"visible_layer\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-8-e35e4611f330>:1: read_data_sets (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use alternatives such as official/mnist/dataset.py from tensorflow/models.\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/mnist.py:260: maybe_download (from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please write your own downloading logic.\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/base.py:252: _internal_retry.<locals>.wrap.<locals>.wrapped_fn (from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use urllib or similar directly.\n",
            "Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/mnist.py:262: extract_images (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use tf.data to implement this functionality.\n",
            "Extracting MNIST_data/train-images-idx3-ubyte.gz\n",
            "Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/mnist.py:267: extract_labels (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use tf.data to implement this functionality.\n",
            "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/mnist.py:110: dense_to_one_hot (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use tf.one_hot on tensors.\n",
            "Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.\n",
            "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n",
            "Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.\n",
            "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n",
            "WARNING:tensorflow:From /tensorflow-1.15.2/python3.6/tensorflow_core/contrib/learn/python/learn/datasets/mnist.py:290: DataSet.__init__ (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use alternatives such as official/mnist/dataset.py from tensorflow/models.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-nzcrMQ4GfHG",
        "colab_type": "code",
        "outputId": "42dbdffb-bc42-47bf-8d90-ac7a832e52ba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 87
        }
      },
      "source": [
        "k=1\n",
        "train_op = rbm_model.update(v, K=k)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-6-49867ba2d0aa>:90: calling reduce_sum_v1 (from tensorflow.python.ops.math_ops) with keep_dims is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "keep_dims is deprecated, use keepdims instead\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "idGbH6boGogx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "energy = rbm_model.free_energy(v=v)\n",
        "tf.summary.scalar('free_energy', tf.reduce_mean(energy))\n",
        "\n",
        "# Merge summaries for Tensorboard visualization\n",
        "summary = tf.summary.merge_all()\n",
        "\n",
        "# Save & restore all the variables.\n",
        "saver = tf.train.Saver({\"weights\":rbm_model.W, \"visible_bias\":rbm_model.vb, \"hidden_bias\":rbm_model.hb})"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wqu_v120GvTD",
        "colab_type": "code",
        "outputId": "1b6a9073-6859-47ba-845c-3a78f0abdaf2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 118
        }
      },
      "source": [
        "\n",
        "# Create session.\n",
        "sess = tf.InteractiveSession()\n",
        "sess.run(tf.global_variables_initializer())\n",
        "test_writer = tf.summary.FileWriter(os.getcwd()+'/RBM_logs/MNIST_Bernoulli',sess.graph)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Issue encountered when serializing trainable_variables.\n",
            "Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.\n",
            "tf.float32 has type DType, but expected one of: int, long, bool\n",
            "WARNING:tensorflow:Issue encountered when serializing variables.\n",
            "Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.\n",
            "tf.float32 has type DType, but expected one of: int, long, bool\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2yjs6_izGzcQ",
        "colab_type": "code",
        "outputId": "ce2e5408-b9de-4828-ed3b-c91d28d1038f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 319
        }
      },
      "source": [
        "\n",
        "for epoch in tqdm(range(rbm_model.epochs)):\n",
        "    if epoch % 100 == 0:\n",
        "        result = sess.run([rbm_model.rec_error, summary], feed_dict = {v: mnist.test.images})\n",
        "        if epoch % 1000 == 0: print(\"Reconstruction error at step {}: {:.3f}\".format(epoch, result[0]))\n",
        "        test_writer.add_summary(result[1],epoch)\n",
        "\n",
        "    batchXs, batchYs = mnist.train.next_batch(32)\n",
        "    sess.run(train_op, feed_dict = {v: batchXs})\n",
        "\n",
        "save_path = saver.save(sess, os.getcwd()+\"/saved_model/actor.ckpt\")\n",
        "print(\"Saving model at: %s\"%(save_path))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "  0%|          | 1/10000 [00:00<1:26:00,  1.94it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 0: 0.341\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 11%|█         | 1059/10000 [00:05<00:49, 179.64it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 1000: 0.024\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 21%|██        | 2054/10000 [00:11<00:42, 185.07it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 2000: 0.020\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 30%|███       | 3043/10000 [00:16<00:37, 184.19it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 3000: 0.019\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 41%|████      | 4065/10000 [00:21<00:32, 184.85it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 4000: 0.017\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5041/10000 [00:26<00:26, 186.95it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 5000: 0.017\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 60%|██████    | 6039/10000 [00:31<00:21, 187.66it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 6000: 0.016\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 70%|███████   | 7038/10000 [00:37<00:17, 171.71it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 7000: 0.015\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 80%|████████  | 8049/10000 [00:42<00:10, 184.13it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 8000: 0.015\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            " 91%|█████████ | 9053/10000 [00:47<00:05, 184.40it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Reconstruction error at step 9000: 0.014\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 10000/10000 [00:52<00:00, 189.90it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Issue encountered when serializing trainable_variables.\n",
            "Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.\n",
            "tf.float32 has type DType, but expected one of: int, long, bool\n",
            "WARNING:tensorflow:Issue encountered when serializing variables.\n",
            "Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.\n",
            "tf.float32 has type DType, but expected one of: int, long, bool\n",
            "Saving model at: /content/saved_model/actor.ckpt\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fsip3itJG3k1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_features(ft_map):\n",
        "    \"\"\" Plot MNIST features in a table.\"\"\"\n",
        "    len_ = int(np.sqrt(rbm_model.n_hidden))\n",
        "    fig = plt.figure()\n",
        "    features = [ft_map[k] for k in range(rbm_model.n_hidden)]\n",
        "    for x in range(len_):\n",
        "        for y in range(len_):\n",
        "            ax = fig.add_subplot(len_, len_, len_*y+x+1)\n",
        "            ax.matshow(features[len_*y+x].reshape((28,28)), cmap = plt.cm.gray)\n",
        "            plt.xticks(np.array([]))\n",
        "            plt.yticks(np.array([]))\n",
        "    plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hYnUQ0j7HR9O",
        "colab_type": "code",
        "outputId": "3223ec78-05e6-4334-d2a9-5b50f777d5f0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 255
        }
      },
      "source": [
        "\n",
        "ft_map = sess.run(rbm_model.get_feature_map())\n",
        "plot_features(ft_map)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVgAAADuCAYAAAB8iB+cAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 144 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-17vULSIHWDy",
        "colab_type": "code",
        "outputId": "b3ed5738-ee3a-44ca-a867-4a4e2c35bd44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 415
        }
      },
      "source": [
        "def show_digit(original, fake):\n",
        "    f, (ax1, ax2) = plt.subplots(1, 2, sharey=True)\n",
        "    ax1.imshow(original.reshape((28,28)),cmap = plt.cm.gray)\n",
        "    ax1.set_title('Original digit')\n",
        "    ax2.imshow(fake.reshape((28,28)),cmap = plt.cm.gray)\n",
        "    ax2.set_title('Sampled digit')\n",
        "    plt.show()\n",
        "\n",
        "train_image = mnist.train.images[1] # from train distribution\n",
        "sample_train = rbm_model.gibbs(1, v=v).eval(session=sess, feed_dict={v: train_image.reshape(1,-1).astype(np.float32)})  \n",
        "show_digit(original=train_image, fake=sample_train)  \n",
        "\n",
        "test_image = mnist.test.images[1] # from test distribution\n",
        "sample_test = rbm_model.gibbs(1, v=v).eval(session=sess, feed_dict={v: test_image.reshape(1,-1).astype(np.float32)})    \n",
        "show_digit(original=test_image, fake=sample_test)  \n",
        "\n",
        "tf.reset_default_graph()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAADHCAYAAAAAoQhGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAADHCAYAAAAAoQhGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S46LvD4BHcAF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}