[12] Convolutional Neural Network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import mkl\n",
    "\n",
    "mkl.set_num_threads(2)\n",
    "np.random.seed(1234)\n",
    "\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.figure_format = 'retina'\n",
    "plt.rcParams[\"figure.figsize\"] = [16, 9]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Handy utility functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def zeros(*dims):\n",
    "    return np.zeros(shape=tuple(dims), dtype=np.float32)\n",
    "\n",
    "def rand(*dims):\n",
    "    return np.random.rand(*dims).astype(np.float32)\n",
    "\n",
    "def chunks(seq, size):\n",
    "    return (seq[pos:pos + size] for pos in range(0, len(seq), size))\n",
    "\n",
    "def as_matrix(vector):\n",
    "    return np.reshape(vector, (-1, 1))\n",
    "\n",
    "def one_hot_encode(labels):\n",
    "    one_hot = zeros(labels.shape[0], np.max(labels) + 1) \n",
    "    one_hot[np.arange(labels.shape[0]), labels] = 1\n",
    "    return one_hot.astype(np.float32)\n",
    "\n",
    "def tiles(examples):\n",
    "    rows_count = examples.shape[0]\n",
    "    cols_count = examples.shape[1]\n",
    "    tile_height = examples.shape[2]\n",
    "    tile_width = examples.shape[3]\n",
    "\n",
    "    space_between_tiles = 2\n",
    "    img_matrix = np.empty(shape=(rows_count * (tile_height + space_between_tiles) - space_between_tiles,  \n",
    "                                 cols_count * (tile_width + space_between_tiles) - space_between_tiles))\n",
    "    img_matrix.fill(np.nan)\n",
    "\n",
    "    for r in range(rows_count):\n",
    "        for c in range(cols_count):\n",
    "            x_0 = r * (tile_height + space_between_tiles)\n",
    "            y_0 = c * (tile_width + space_between_tiles)\n",
    "            ex_min = np.min(examples[r, c])\n",
    "            ex_max = np.max(examples[r, c])\n",
    "            img_matrix[x_0:x_0 + tile_height, y_0:y_0 + tile_width] = (examples[r, c] - ex_min) / (ex_max - ex_min)\n",
    "    \n",
    "    plt.matshow(img_matrix, cmap='gray', interpolation='none')\n",
    "    plt.axis('off')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Activation functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def relu(batch):\n",
    "    return np.maximum(0.0, batch)\n",
    "\n",
    "def relu_derivative(batch):\n",
    "    return (batch > 0).astype(np.float32)\n",
    "\n",
    "def softmax(batch):\n",
    "    m = as_matrix(np.max(batch, axis=1))\n",
    "    exponents = np.exp(batch - m)\n",
    "    return exponents / as_matrix(np.sum(exponents, axis=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x574.396 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 459,
       "width": 907
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import mnist\n",
    "digits = np.reshape(mnist.train_images()[:12*24], newshape=(12, 24, 28, 28))\n",
    "tiles(digits)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Network layers\n",
    "\n",
    "We define three types of layers: densely connected layer (```Dense```), convolutional layer (```Conv2D```) and non-linearity (```Nonlinear```). Each layer supports two operations:\n",
    "\n",
    "- signal propagation: ```forward```\n",
    "- error backpropagation: ```backward```\n",
    "\n",
    "```backward``` operation backpropagates the delta and updates layer weights and biases.\n",
    "\n",
    "Weights are initialized with \"variance scaling\" method (also known as \"He initialization\"). This initialization was proposed together with residual networks and is appropriate for networks with ReLU activations. Source: https://arxiv.org/pdf/1502.01852v1.pdf\n",
    "\n",
    "**Note that in this lab we have separate vectors for biases.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\npopularne podejscie w frameworkach, w warstwahc faktcznych czesto nie fefiniuje sie funkcji aktywacji.\\npo takiej pustej warstwe dodaje sie warstwe Nonlinear już z zadaną funkcja aktywacji.\\nTo jest czytelniejsze.\\n'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "popularne podejscie w frameworkach, w warstwahc faktcznych czesto nie fefiniuje sie funkcji aktywacji.\n",
    "po takiej pustej warstwe dodaje sie warstwe Nonlinear już z zadaną funkcja aktywacji.\n",
    "To jest czytelniejsze.\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Non-linearity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Nonlinear:\n",
    "    def __init__(self, activation_fun, d_activation_fun):\n",
    "        self.activation_fun = activation_fun\n",
    "        self.d_activation_fun = d_activation_fun\n",
    "        \n",
    "        self.visible = None\n",
    "    \n",
    "    def reset(self):\n",
    "        pass\n",
    "    \n",
    "    def forward(self, batch):\n",
    "        self.visible = batch\n",
    "        return self.activation_fun(batch)\n",
    "    \n",
    "    def backward(self, deltas):\n",
    "        return deltas * self.d_activation_fun(self.visible)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Densly connected layer\n",
    "\n",
    "Implement missing parts in the ```backward``` method. Remember that the biases are kept in a separate vector."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Dense:\n",
    "    def __init__(self, visible_size, hidden_size, learning_rate, momentum):\n",
    "        self.visible_size = visible_size\n",
    "        self.hidden_size = hidden_size\n",
    "                \n",
    "        self.learning_rate = learning_rate\n",
    "        self.momentum = momentum\n",
    "        \n",
    "        self.reset()\n",
    "    \n",
    "    def reset(self):\n",
    "        self.W = np.random.normal(scale=np.sqrt(2.0 / self.visible_size),\n",
    "                                  size=(self.visible_size, self.hidden_size)).astype(np.float32)\n",
    "        self.B = zeros(1, self.hidden_size)\n",
    "        \n",
    "        self.in_shape = None\n",
    "        \n",
    "        self.visible = None\n",
    "        self.hidden  = None\n",
    "\n",
    "        self.MW = zeros(self.visible_size, self.hidden_size)\n",
    "        self.MB = zeros(1, self.hidden_size)\n",
    "\n",
    "    def forward(self, batch):\n",
    "        self.in_shape = batch.shape\n",
    "        self.visible = np.reshape(batch, (batch.shape[0], -1))\n",
    "        \n",
    "        self.hidden = self.visible @ self.W + self.B\n",
    "        return self.hidden\n",
    "    \n",
    "    def backward(self, deltas):\n",
    "        observations_count = self.visible.shape[0]\n",
    "        \n",
    "        prev_deltas = deltas @ self.W.T\n",
    "        prev_deltas = prev_deltas.reshape(self.in_shape)\n",
    "        \n",
    "\n",
    "        # Calculate the gradient for the weights and biases. Store the result\n",
    "        # in `grad_w` and `grad_b`, respectively.\n",
    "        #\n",
    "        # This is a standard densly connected layer, so take hints from the backprop\n",
    "        # implementation in previous labs. Note, however, that we now have a separate\n",
    "        # vector for biases. This means that you need to deduce an explicit equation\n",
    "        # for the gradient with respect to the biases (consider how were the biases\n",
    "        # updated in the previous backprop implementation).\n",
    "        #\n",
    "        # Inputs to the current layer are stored in `self.visible`\n",
    "        # Deltas for the current layer are stored in `deltas`.\n",
    "        \n",
    "        # gradient w dla warstw konwolucyjnych, gradient b dla biasów.\n",
    "        grad_w = self.visible.T @ deltas # visible is transposed because of backward operation.\n",
    "        grad_b = 1 * np.sum(deltas) \n",
    "        # in previous labs, we added special '1' into vector of\n",
    "        # observations to enable multiplication.\n",
    "        \n",
    "        \n",
    "        grad_b = np.expand_dims(grad_b, axis=0)\n",
    "        \n",
    "        self.MW = self.momentum * self.MW - \\\n",
    "                  self.learning_rate / observations_count * grad_w\n",
    "        self.MB = self.momentum * self.MB - \\\n",
    "                  self.learning_rate / observations_count * grad_b\n",
    "        \n",
    "        self.W += self.MW\n",
    "        self.B += self.MB\n",
    "        \n",
    "        return prev_deltas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convolutional layers\n",
    "\n",
    "We implement the convolution operation as a simple matrix multiplication. Essentially, convolution is a dot product between the kernel and the receptive fields (local regions of the input image). Therefore:\n",
    "- We can rearrange the input volume into a matrix $\\mathbf{X}$, in which each column is a receptive field reshaped to a vector. The number of columns in this matrix is equal to the number of receptive fields in an image multiplied by the number of input images (batch size).\n",
    "- Next, we can reshape the convolution kernel into a vector. Note, however, that a convolutional layer has many kernels: the number of kernels is equal to the number of output channels. We can assemble these kernels into a matrix $\\mathbf{K}$, in which each row is one kernel (reshaped to a vector).\n",
    "- Once we have matrices $\\mathbf{X}$ and $\\mathbf{K}$, calculating the convolution is as simple as taking a matrix multiplication $\\mathbf{KX}$.\n",
    "\n",
    "A very important property of this implementation is that the convolution is expressed as a matrix multiplication. This is the same basic operation that is used in densely connected layers. This means that backpropagation equations from a simple MLPs also holds for this implementation of a convolutional layer. Matrix $\\mathbf{X}$ is the equivalent of the visible layer and matrix $\\mathbf{K}$ is the equivalent of the weight matrix.\n",
    "\n",
    "Important notes:\n",
    "- The rearrangements of the input volume into a matrix is implemented by ```im2col``` function.\n",
    "- During backpropagation we need to perform an operation inverse to ```im2col```. Specifically, we need to “collapse” deltas from a matrix into a volume with shape equal to the shape of the input. This inverse operation is implemented by ```col2im``` function.\n",
    "- Kernels in ```Conv2D``` are always kept as a matrix, where each row is a kernel for one output channel.\n",
    "- Convolution layer has no activation function (or, equivalently, has an identify activation function). Nonlinear activations are implemented by a separate layer (```Nonlinear``` layer). Therefore, we can disregard activation function (and its derivative) when implementing error backpropagation in ```Conv2D```.\n",
    "- We have a separate vector $\\mathbf{B}$ for biases. Our expression for convolution is therefore: $\\mathbf{KX} + \\mathbf{B}$. Note that biases are shared between receptive fields (vector $\\mathbf{B}$ is broadcasted to the shape of $\\mathbf{KX}$). See implementation of ```Dense``` layer for hints about proper implementation of biases in forward/backward pass.\n",
    "\n",
    "You can find more information about implementing convolution via ```im2col```/```col2im``` here:\n",
    "* http://cs231n.stanford.edu/slides/2016/winter1516_lecture11.pdf\n",
    "* http://cs231n.github.io/convolutional-networks/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### im2col and col2im"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def col_indices(shape, kernel_size, strides):\n",
    "    kh, kw = kernel_size\n",
    "    sh, sw = strides\n",
    "    \n",
    "    n, c, h, w = shape\n",
    "    \n",
    "    col_extent = w - kw + 1\n",
    "    row_extent = h - kh + 1\n",
    "    \n",
    "    batch_idx = np.arange(n)[:, None, None] * c * h * w\n",
    "    \n",
    "    start_idx = np.arange(kh)[None, :, None] * w + np.arange(kw)\n",
    "    \n",
    "    depth_idx = h * w * np.arange(c)\n",
    "    start_idx = (depth_idx[None, :, None] + start_idx.ravel())\n",
    "    start_idx = start_idx.reshape((-1, kh, kw))\n",
    "    \n",
    "    offset_idx = np.arange(row_extent)[None, :, None] * w + np.arange(col_extent)\n",
    "    \n",
    "    idx = (batch_idx + \n",
    "           start_idx.ravel()[None, :, None] + \n",
    "           offset_idx[:, ::sh, ::sw].ravel())\n",
    "    \n",
    "    return idx\n",
    "\n",
    "def im2col(batch, kernel_size, strides, cached_indices=None):\n",
    "    if cached_indices is not None and batch.shape[0] == cached_indices.shape[0]:\n",
    "        idx = cached_indices\n",
    "    else:\n",
    "        idx = col_indices(batch.shape, kernel_size, strides)\n",
    "    \n",
    "    cols = np.take(batch, idx).transpose(1, 0, 2)\n",
    "    cols = np.reshape(cols, (cols.shape[0], -1))\n",
    "    return cols, idx\n",
    "\n",
    "def col2im(shape, cols, kernel_size, strides, cached_indices=None):\n",
    "    n, c, h, w = shape\n",
    "    kh, kw = kernel_size\n",
    "    \n",
    "    if cached_indices is not None and cols.shape == cached_indices.shape:\n",
    "        idx = cached_indices\n",
    "    else:\n",
    "        idx = col_indices(shape, kernel_size, strides).transpose(1, 0, 2).reshape(kh*kw*c, -1)\n",
    "    \n",
    "    im = zeros(*shape).reshape(-1)\n",
    "    for idxs, vals in zip(idx, cols):\n",
    "        im[idxs] += vals\n",
    "    \n",
    "    return im.reshape(*shape), idx"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2D convolution\n",
    "\n",
    "Implement missing parts in the ```forward``` and ```backward``` methods.  Remember that the biases are kept in a separate vector."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Conv2D:\n",
    "    def __init__(self,\n",
    "                 in_channels,\n",
    "                 out_channels,\n",
    "                 kernel_size,\n",
    "                 strides,\n",
    "                 padding,\n",
    "                 learning_rate,\n",
    "                 momentum,\n",
    "                 cache_col_indices=True):\n",
    "        self.in_channels = in_channels\n",
    "        self.out_channels = out_channels\n",
    "        self.kernel_size = kernel_size\n",
    "        self.strides = strides\n",
    "        self.padding = padding\n",
    "        \n",
    "        kh, kw = self.kernel_size\n",
    "        self.klen = kh * kw * self.in_channels\n",
    "        \n",
    "        self.learning_rate = learning_rate\n",
    "        self.momentum = momentum\n",
    "        \n",
    "        self.cache_col_indices = cache_col_indices\n",
    "                    \n",
    "        self.reset()\n",
    "    \n",
    "    def reset(self):\n",
    "        self.K = np.random.normal(scale=np.sqrt(2.0 / self.klen),\n",
    "                                  size=(self.out_channels, self.klen)).astype(np.float32)\n",
    "        \n",
    "        self.B = zeros(self.out_channels, 1) + 1e-2 # Initialize biases to small positive values (for ReLU)\n",
    "        \n",
    "        self.padded_shape = None\n",
    "        self.cols         = None\n",
    "        \n",
    "        self.i2c_col_indeces = None\n",
    "        self.c2i_col_indeces = None\n",
    "        \n",
    "        self.MK = zeros(self.out_channels, self.klen)\n",
    "        self.MB = zeros(self.out_channels, 1)\n",
    "    \n",
    "    def pad(self, batch):\n",
    "        ph, pw = self.padding\n",
    "        if ph == 0 and pw == 0:\n",
    "            self.padded_shape = batch.shape\n",
    "            return batch\n",
    "\n",
    "        n, c, h, w = batch.shape        \n",
    "        batch_padded = zeros(n, c, h + 2 * ph, w + 2 * pw)\n",
    "        batch_padded[:, :, ph:(ph + h), pw:(pw + w)] = batch\n",
    "        \n",
    "        self.padded_shape = batch_padded.shape\n",
    "        \n",
    "        return batch_padded\n",
    "    \n",
    "    def unpad(self, deltas_padded):\n",
    "        ph, pw = self.padding\n",
    "        \n",
    "        deltas = deltas_padded[:, :, ph:-ph, :] if ph > 0 else deltas_padded\n",
    "        deltas = deltas[:, :, :, pw:-pw] if pw > 0 else deltas\n",
    "        \n",
    "        return deltas\n",
    "    \n",
    "    def output_size(self):\n",
    "        kh, kw = self.kernel_size\n",
    "        sh, sw = self.strides\n",
    "        _, _, padded_h, padded_w = self.padded_shape\n",
    "        \n",
    "        out_h = (padded_h - kh) / sh + 1\n",
    "        out_w = (padded_w - kw) / sw + 1\n",
    "        return int(out_h), int(out_w)\n",
    "    \n",
    "    def forward(self, batch):\n",
    "        \n",
    "        batch_padded = self.pad(batch)\n",
    "        out_h, out_w = self.output_size()\n",
    "\n",
    "        self.cols, i2c_col_indeces = im2col(batch_padded, self.kernel_size, self.strides, self.i2c_col_indeces)\n",
    "        if self.cache_col_indices:\n",
    "            self.i2c_col_indeces = i2c_col_indeces\n",
    "\n",
    "\n",
    "        # Calculate the convolution between the rearranged input volume and the convolution kernels.\n",
    "        # Add biases to the result. Store the final result in 'output' matrix.\n",
    "        #\n",
    "        # - the rearranged input volume is stored in 'self.cols'\n",
    "        # - the convolution kernels are stored in 'self.K'\n",
    "        # - the biases are stored in 'self.B'\n",
    "        \n",
    "        output = self.K @ self.cols + self.B # self.K @ self.cols to dopowiedniek KX\n",
    "        \n",
    "        # Using 'reshape' and 'transpose' NumPy operations reshape the 'output' matrix to the\n",
    "        # shape of the output volume.\n",
    "        #\n",
    "        # Currently 'output' matrix has the shape:\n",
    "        #   self.out_channels x (batch_size * out_h * out_w)\n",
    "        # where the order of elements in the second dimension is:\n",
    "        #   examples (batch_size) -> height (out_h) -> width (out_w).\n",
    "        #\n",
    "        # The output volume should have the shape:\n",
    "        #   batch_size x self.out_channels x out_h x out_w\n",
    "        output = output.reshape(\n",
    "            (self.out_channels, len(batch_padded), out_h, out_w)\n",
    "        )\n",
    "        \n",
    "        output = np.transpose(output, (1, 0, 2, 3))\n",
    "\n",
    "        return output\n",
    "    \n",
    "    def backward(self, deltas):\n",
    "        observations_count = self.padded_shape[0]\n",
    "        \n",
    "        deltas = deltas.transpose(1, 0, 2, 3)\n",
    "        deltas = deltas.reshape(self.out_channels, -1)\n",
    "        \n",
    "        \n",
    "        #raise NotImplementedError('Backpropagation in Conv2D is not implemented.')\n",
    "        \n",
    "        # Calculate the deltas for the previous layer (i.e backpropagate the errors).\n",
    "        # Store the result in 'prev_deltas'.\n",
    "        #\n",
    "        # Deltas for the current layer are stored in `deltas`.\n",
    "        \n",
    "        prev_deltas = self.K.T @ deltas\n",
    "        \n",
    "        \n",
    "        prev_deltas, c2i_col_indeces = col2im(self.padded_shape,\n",
    "                                              prev_deltas,\n",
    "                                              self.kernel_size,\n",
    "                                              self.strides,\n",
    "                                              self.c2i_col_indeces)\n",
    "        if self.cache_col_indices:\n",
    "            self.c2i_col_indeces = c2i_col_indeces        \n",
    "        \n",
    "        prev_deltas = self.unpad(prev_deltas)\n",
    "        \n",
    "\n",
    "        #raise NotImplementedError('Gradient calculation in Conv2D is not implemented.')\n",
    "        \n",
    "        # Calculate the gradient for the kernels and biases. Store the result\n",
    "        # in `grad_k` and `grad_b`, respectively.\n",
    "\n",
    "        grad_k = deltas @ self.cols.T\n",
    "        grad_b = 1 * np.sum(deltas, axis=1)\n",
    "        \n",
    "        \n",
    "        grad_b = np.expand_dims(grad_b, axis=1)\n",
    "\n",
    "        self.MK = self.momentum * self.MK - \\\n",
    "                  self.learning_rate / observations_count * grad_k\n",
    "        self.MB = self.momentum * self.MB - \\\n",
    "                  self.learning_rate / observations_count * grad_b\n",
    "        \n",
    "        self.K += self.MK\n",
    "        self.B += self.MB\n",
    "        \n",
    "        return prev_deltas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Error backpropagation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def forward_pass(mlp, batch):\n",
    "    for layer in mlp:\n",
    "        batch = layer.forward(batch)\n",
    "        \n",
    "    return batch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def error_backpropagate(mlp, deltas):\n",
    "    for layer in reversed(mlp):\n",
    "        deltas = layer.backward(deltas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MLP training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_mlp(mlp, dataset, labels, batch_size):\n",
    "    batches_limit = dataset.shape[0] // batch_size\n",
    "    \n",
    "    batched_data = chunks(dataset, batch_size)\n",
    "    batched_labels = chunks(labels, batch_size)\n",
    "    \n",
    "    for batch_idx, (batch, batch_labels) in enumerate(zip(batched_data, batched_labels)):\n",
    "\n",
    "        y = forward_pass(mlp, batch)\n",
    "        \n",
    "        deltas = y - batch_labels\n",
    "        error_backpropagate(mlp[0:-1], deltas)\n",
    "        \n",
    "        if batch_idx % round(batches_limit / 40) == 0: print(\"#\", end=\"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST digits classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classify(mlp, batch):\n",
    "    probabilities = np.concatenate([forward_pass(mlp, mini_batch) for mini_batch in chunks(batch, 128)])\n",
    "    return np.argmax(probabilities, axis=1)\n",
    "\n",
    "def run_training(mlp, train_set, train_labels,\n",
    "                 validation_set, validation_labels,\n",
    "                 batch_size, epochs_count):\n",
    "    \n",
    "    for epoch in range(epochs_count):\n",
    "        print(\"Epoch {}:\".format(epoch+1),  end=\"\\t\")\n",
    "\n",
    "        if epoch == 5:\n",
    "            for layer in mlp:\n",
    "                layer.momentum = 0.9\n",
    "                \n",
    "        start_time = time.time()\n",
    "        train_mlp(mlp, train_set, train_labels, batch_size)\n",
    "        elapsed = time.time() - start_time\n",
    "\n",
    "        predictions = classify(mlp, validation_set)\n",
    "        accuracy = 100.0 * np.sum(predictions == validation_labels) / predictions.shape[0]\n",
    "        print(\"\\telapsed: {0:>2.2f}s, accuracy: {1:>2.2f}\".format(elapsed, accuracy))\n",
    "\n",
    "    print(\"Training finished!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "DATASET_SIZE = 10000 # 60000 for whole dataset\n",
    "DIGIT_SIZE = 28\n",
    "\n",
    "##### Train set #####\n",
    "\n",
    "mnist_train_images = mnist.train_images().astype(np.float32) / 255.0\n",
    "mnist_train_labels = mnist.train_labels()\n",
    "\n",
    "order = np.random.permutation(len(mnist_train_images))\n",
    "mnist_train_images = mnist_train_images[order]\n",
    "mnist_train_labels = mnist_train_labels[order]\n",
    "\n",
    "mnist_train_images = mnist_train_images[:DATASET_SIZE]\n",
    "mnist_train_images = mnist_train_images.reshape(-1, 1, DIGIT_SIZE, DIGIT_SIZE)\n",
    "\n",
    "mnist_train_labels = mnist_train_labels[:DATASET_SIZE]\n",
    "mnist_train_labels = one_hot_encode(mnist_train_labels)\n",
    "\n",
    "##### Test set #####\n",
    "\n",
    "mnist_test_images = mnist.test_images().astype(np.float32) / 255.0\n",
    "mnist_test_images = mnist_test_images.reshape(-1, 1, DIGIT_SIZE, DIGIT_SIZE)\n",
    "\n",
    "mnist_test_labels = mnist.test_labels()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1:\t########################################\telapsed: 3.79s, accuracy: 58.52\n",
      "Epoch 2:\t########################################\telapsed: 4.09s, accuracy: 72.90\n",
      "Epoch 3:\t########################################\telapsed: 4.13s, accuracy: 82.69\n",
      "Epoch 4:\t########################################\telapsed: 3.84s, accuracy: 86.20\n",
      "Epoch 5:\t########################################\telapsed: 3.69s, accuracy: 87.96\n",
      "Epoch 6:\t########################################\telapsed: 3.67s, accuracy: 93.87\n",
      "Epoch 7:\t########################################\telapsed: 3.80s, accuracy: 94.44\n",
      "Epoch 8:\t########################################\telapsed: 4.39s, accuracy: 94.78\n",
      "Epoch 9:\t########################################\telapsed: 3.78s, accuracy: 95.55\n",
      "Epoch 10:\t########################################\telapsed: 4.55s, accuracy: 95.69\n",
      "Epoch 11:\t########################################\telapsed: 4.43s, accuracy: 95.97\n",
      "Epoch 12:\t########################################\telapsed: 4.51s, accuracy: 96.15\n",
      "Epoch 13:\t########################################\telapsed: 4.28s, accuracy: 96.00\n",
      "Epoch 14:\t########################################\telapsed: 4.19s, accuracy: 95.40\n",
      "Epoch 15:\t########################################\telapsed: 3.92s, accuracy: 95.68\n",
      "Epoch 16:\t########################################\telapsed: 5.91s, accuracy: 95.66\n",
      "Epoch 17:\t########################################\telapsed: 8.20s, accuracy: 95.44\n",
      "Epoch 18:\t########################################\telapsed: 8.31s, accuracy: 95.72\n",
      "Epoch 19:\t########################################\telapsed: 6.98s, accuracy: 95.45\n",
      "Epoch 20:\t########################################\telapsed: 3.71s, accuracy: 96.37\n",
      "Epoch 21:\t########################################\telapsed: 5.36s, accuracy: 96.61\n",
      "Epoch 22:\t########################################\telapsed: 8.32s, accuracy: 96.62\n",
      "Epoch 23:\t########################################\telapsed: 12.82s, accuracy: 96.81\n",
      "Epoch 24:\t########################################\telapsed: 10.40s, accuracy: 96.45\n",
      "Epoch 25:\t########################################\telapsed: 8.01s, accuracy: 96.51\n",
      "Epoch 26:\t########################################\telapsed: 5.67s, accuracy: 96.59\n",
      "Epoch 27:\t########################################\telapsed: 10.39s, accuracy: 96.54\n",
      "Epoch 28:\t########################################\telapsed: 8.58s, accuracy: 96.50\n",
      "Epoch 29:\t########################################\telapsed: 8.01s, accuracy: 96.73\n",
      "Epoch 30:\t########################################\telapsed: 7.75s, accuracy: 96.62\n",
      "Epoch 31:\t########################################\telapsed: 7.95s, accuracy: 96.14\n",
      "Epoch 32:\t########################################\telapsed: 9.12s, accuracy: 96.61\n",
      "Epoch 33:\t########################################\telapsed: 10.01s, accuracy: 96.63\n",
      "Epoch 34:\t########################################\telapsed: 10.20s, accuracy: 96.81\n",
      "Epoch 35:\t########################################\telapsed: 8.31s, accuracy: 96.89\n",
      "Epoch 36:\t########################################\telapsed: 8.49s, accuracy: 96.57\n",
      "Epoch 37:\t########################################\telapsed: 7.96s, accuracy: 96.77\n",
      "Epoch 38:\t########################################\telapsed: 8.05s, accuracy: 96.44\n",
      "Epoch 39:\t########################################\telapsed: 7.62s, accuracy: 96.63\n",
      "Epoch 40:\t########################################\telapsed: 7.79s, accuracy: 96.35\n",
      "Epoch 41:\t########################################\telapsed: 7.84s, accuracy: 96.39\n",
      "Epoch 42:\t########################################\telapsed: 7.78s, accuracy: 96.75\n",
      "Epoch 43:\t########################################\telapsed: 8.94s, accuracy: 96.43\n",
      "Epoch 44:\t########################################\telapsed: 8.52s, accuracy: 96.57\n",
      "Epoch 45:\t########################################\telapsed: 8.60s, accuracy: 96.90\n",
      "Epoch 46:\t########################################\telapsed: 8.10s, accuracy: 97.08\n",
      "Epoch 47:\t########################################\telapsed: 7.77s, accuracy: 97.08\n",
      "Epoch 48:\t########################################\telapsed: 7.87s, accuracy: 97.11\n",
      "Epoch 49:\t########################################\telapsed: 7.80s, accuracy: 97.13\n",
      "Epoch 50:\t########################################\telapsed: 7.94s, accuracy: 97.14\n",
      "Training finished!\n"
     ]
    }
   ],
   "source": [
    "BATCH_SIZE = 128\n",
    "EPOCHS_COUNT = 50\n",
    "\n",
    "LEARNING_RATE = 0.01\n",
    "MOMENTUM = 0.5\n",
    "\n",
    "mlp = [\n",
    "    Conv2D(\n",
    "        in_channels=1,\n",
    "        out_channels=2,\n",
    "        kernel_size=(3, 3),\n",
    "        strides=(1, 1),\n",
    "        padding=(1, 1),\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(relu, relu_derivative),\n",
    "    Conv2D(\n",
    "        in_channels=2,\n",
    "        out_channels=2,\n",
    "        kernel_size=(2, 2),\n",
    "        strides=(2, 2),\n",
    "        padding=(0, 0),\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(relu, relu_derivative),\n",
    "    Conv2D(\n",
    "        in_channels=2,\n",
    "        out_channels=4,\n",
    "        kernel_size=(3, 3),\n",
    "        strides=(1, 1),\n",
    "        padding=(1, 1),\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(relu, relu_derivative),\n",
    "    Conv2D(\n",
    "        in_channels=4,\n",
    "        out_channels=4,\n",
    "        kernel_size=(2, 2),\n",
    "        strides=(2, 2),\n",
    "        padding=(0, 0),\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(relu, relu_derivative),\n",
    "    Conv2D(\n",
    "        in_channels=4,\n",
    "        out_channels=32,\n",
    "        kernel_size=(3, 3),\n",
    "        strides=(1, 1),\n",
    "        padding=(1, 1),\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(relu, relu_derivative),\n",
    "    Dense(\n",
    "        visible_size=32 * 7 * 7,\n",
    "        hidden_size=10,\n",
    "        learning_rate=LEARNING_RATE,\n",
    "        momentum=MOMENTUM),\n",
    "    Nonlinear(softmax, None)\n",
    "]\n",
    "\n",
    "run_training(mlp,\n",
    "             mnist_train_images, mnist_train_labels,\n",
    "             mnist_test_images, mnist_test_labels,\n",
    "             BATCH_SIZE, EPOCHS_COUNT)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "NNaDL",
   "language": "python",
   "name": "nnadl"
  },
  "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": 2
}