From 4b4c218c0420b0cc23f25e4d515eb0e94f53e1c2 Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Tue, 3 Aug 2021 22:54:11 -0400 Subject: [PATCH 1/8] add lesson 7 and scripts --- .../7_The_First_Deep_Neural_Network.ipynb | 791 ++++++++++++++++++ scripts/lesson_5_functions.py | 110 +++ scripts/load_casting_data.py | 84 ++ 3 files changed, 985 insertions(+) create mode 100644 notebooks_en/7_The_First_Deep_Neural_Network.ipynb create mode 100644 scripts/lesson_5_functions.py create mode 100644 scripts/load_casting_data.py diff --git a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb new file mode 100644 index 0000000..7a42e46 --- /dev/null +++ b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb @@ -0,0 +1,791 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# From multiple logistic regression to the very first deep neural network" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "In lesson 5, we showed how to apply a logistic regression model to identify defective metal-casting parts. However, the accuracy seems not good enough. This lesson will improve the accuracy by replacing the logistic model with a more advanced model: a fully connected neural network model with one hidden layer. Fully connected neural networks are the cornerstone of deep learning and so-called artificial intelligence." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Before we start" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Before we start, let's import the required libraries:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from autograd import numpy\n", + "from autograd import grad\n", + "from matplotlib import pyplot" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Also, we will use random numbers in this lesson. As this is a teaching material, we want to control the generation of random numbers to make it NOT really random -- you will get the same results as we did when executing this notebook by fixing the random seed:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "numpy.random.seed(0)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "In this lesson, our example application is identifying defective metal-casting parts using computer vision. We already saw this application in lesson 5. To save space, we wrapped the code loading and data normalizing in file `scripts/load_casting_data.py`. Execute the following code to get the data from the wrapper: " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "import sys\n", + "sys.path.insert(0, \"../scripts\")\n", + "from load_casting_data import *" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The above code cell loads the following variables: `res`, `n_ok_total`, `n_ok_train`, `n_ok_val`, `n_ok_test`, `n_def_total`, `n_def_train`, `n_def_val`, `n_def_test`, `images_train`, `images_val`, `images_test`, `labels_train`, `labels_val`, `labels_test`, `mu`, and `sigma`. Though the variable names should be self-explanatory enough, it's still a good idea to revisit lesson 5 for the meaning of these variables." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Also, we will reuse some functions from lesson 5. We wrapped these functions in `scripts/lesson_5_functions.py`. Let's also import these functions:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "from lesson_5_functions import *" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The above code will import `logistic`, `classify`, and `performance`. They are the same code as what we used in lesson 5." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Introduction to fully connected neural networks" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### What are fully connected neural networks?" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "While neural networks in machine learning may be inspired by real neural networks in our bodies, from an engineering perspective, they are just mathematical models and nothing magical. Most of them can be expressed with high-school math. Recall that a mathematical model is a hypothesized relationship between input features $\\mathbf{x}$ and outputs $\\mathbf{y}$. For example, a linear regression model assumes $\\mathbf{y} \\approx \\hat{\\mathbf{y}} = W^\\mathsf{T}\\mathbf{x}+\\mathbf{b}$, and a logistic regression model assumes $\\mathbf{y}\\approx\\hat{\\mathbf{y}}=\\operatorname{logistic}\\left(W^\\mathsf{T}\\mathbf{x}+\\mathbf{b}\\right)$. Neural networks assume a more complicated relationship between $\\mathbf{x}$ and $\\mathbf{y}$." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "There is no unanimous definition of neural networks. However, fully connected neural networks are commonly deemed the most basic type of neural network models among various kinds of neural networks. An input vector $\\mathbf{x}$ goes through several consecutive and interleaved linear and nonlinear transformations in fully connected neural networks before the model returns $\\hat{\\mathbf{y}}$. A linear transformation is something like $W^\\mathsf{T}\\mathbf{x}$, and a nonlinear transformation may be a logistic function, a trigonometric function, or any functions that are not linear." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Given a single input vector $\\mathbf{x}$, a fully connected neural network model returns $\\hat{\\mathbf{y}}$ through: \n", + "\n", + "$$\n", + "\\begin{aligned}\n", + "\\mathbf{z}^1 &= \\sigma_0\\left(\\left(W^0\\right)^\\mathsf{T}\\mathbf{x} + \\mathbf{b}^0\\right) \\\\\n", + "\\mathbf{z}^2 &= \\sigma_1\\left(\\left(W^1\\right)^\\mathsf{T}\\mathbf{z}^1 + \\mathbf{b}^1\\right) \\\\\n", + "&\\vdots \\\\\n", + "\\hat{\\mathbf{y}} &= \\sigma_L\\left(\\left(W^L\\right)^\\mathsf{T}\\mathbf{z}^L + \\mathbf{b}^L\\right)\n", + "\\end{aligned}\n", + "$$\n", + "\n", + "Or, if we prefer an iterative expression:\n", + "\n", + "$$\n", + "\\begin{aligned}\n", + "\\mathbf{z}^0 &\\equiv \\mathbf{x} \\\\\n", + "\\mathbf{z}^{i+1} &= \\sigma_{i}\\left(\\left(W^i\\right)^\\mathsf{T}\\mathbf{z}^{i} + \\mathbf{b}^{i}\\right)\\text{ for }i=0,\\dots,L \\\\\n", + "\\hat{\\mathbf{y}} &\\equiv \\mathbf{z}^{L+1}\n", + "\\end{aligned}\n", + "$$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We use bold-face symbols because these variables are vectors and matrices, as seen in lesson 6. The math equations of the linear and logistic regression models in lessons 1 to 5 are special cases of the vector-matrix form. Let's take a look at these symbols:" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "- Linear transformations and model parameters: $W^0$, $\\mathbf{b}^0$, $\\dots$, $W^L$, $\\mathbf{b}^L$\n", + " \n", + " $W^0$, $\\mathbf{b}^0$, $\\dots$, $W^L$, $\\mathbf{b}^L$ are parameter matrices and vectors. They are responsible for linear transformations, just like in previous lessons. The difference in this lesson is that now we have more than one linear transformation. We will determine the values of these parameters using gradient descent in this lesson.\n", + " \n", + " Note that the superscript of the parameter matrices and vectors starts from $0$. Usually, in math expression, we start indices with $1$. For example, the first element in a vector $\\mathbf{x}$ is $x_1$. However, we begin with $0$ for $W$ and $\\mathbf{b}$ because we want to address how many ***additional*** transformations we have in the model besides the last transformation, i.e., the one that returns $\\mathbf{y}$. This information is carried by the variable $L$." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "- Nonlinear transformations: $\\sigma_0$, $\\sigma_1$, $\\dots$, $\\sigma_{L}$\n", + " \n", + " $\\sigma_0$, $\\sigma_1$, $\\dots$, $\\sigma_{L}$ are choosen nonlinear functions. It's not uncommon to use the same function for all of them, i.e., $\\sigma_0 = \\sigma_1 = \\dots = \\sigma_{L} = \\sigma$. In addition, they are often element-wise functions. That says, if we give a vector $\\mathbf{g}$ to $\\sigma_0$, then\n", + " \n", + " $$\n", + " \\sigma_0\\left(\\mathbf{g}\\right)=\n", + " \\sigma_0\\left(\\begin{bmatrix}g_1 \\\\ g_2 \\\\ \\vdots \\end{bmatrix} \\right)=\n", + " \\begin{bmatrix}\\sigma_0(g_1) \\\\ \\sigma_0(g_2) \\\\ \\vdots \\end{bmatrix}\n", + " $$" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "- Intermediate results: $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$\n", + " \n", + " $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$ are intermediate results. In rare situations, we can explain the meaning of these intermediate results. For example, if we are modeling a natural phenomenon, these results may represent some physical mechanisms or properties. However, in most applications, these intermediate results are just the side products of the calculations and do not carry any meaning. " + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "- Number of linear-nonlinear-transformation pairs: $L$\n", + " \n", + " $L$ is a user-defined parameter that controls how many linear-nonlinear transformation pairs before the model returns the output $\\mathbf{y}$. We use $L$ to adjust the complexity of a fully connected neural network model: larger $L$ makes our model more complicated. And a more complicated neural network can model a more sophisticated relationship between $\\mathbf{x}$ and $\\mathbf{y}$.\n", + " \n", + " As mentioned previously, $L$ denotes how many additional transformations we have in the model before the one returning the final result. Alternatively, we can also treat $L$ as how many intermediate results we have in the model." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "- Lengths of intermediate vectors: $n_{z^1}$, $n_{z^2}$, $\\dots$, $n_{z^L}$\n", + " \n", + " $n_{z^1}$, $n_{z^2}$, $\\dots$, $n_{z^L}$ are not visible in the above matrix-vector form of the model. They represent the length of the intermediate vectors $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. These variables are user-defined and also control the complexity of models." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "If we choose $\\sigma_0(\\mathbf{g}) = \\sigma_1(\\mathbf{g}) = \\dots = \\sigma_{L}(\\mathbf{g}) = \\operatorname{logistic}(\\mathbf{g}) = \\frac{1}{1+\\mathrm{e}^{-\\mathbf{g}}}$, then a fully connected neural network is a chain of several logistic regression models. In fact, ***a logistic regression model is a fully connected neural network*** with $L=0$." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Our first neural network" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We will use the following configuration for the fully-connected neural network in this lesson:\n", + "\n", + "1. $L=1$,\n", + "2. $\\sigma_0(\\mathbf{g})=\\sigma_1(\\mathbf{g})=\\operatorname{logistic}(\\mathbf{g})$, and\n", + "3. $n_{z^1}=64$, meaning $\\mathbf{z}^1=\\begin{bmatrix}z_1^1 & z_2^1 & \\cdots & z_{64}^1 \\end{bmatrix}^\\mathsf{T}$\n", + "\n", + "Also, each of our images has a total of $128\\times 128=16384$ pixels (revisit lesson 5 if you don't remember), so $\\mathbf{x}=\\begin{bmatrix}x_1 & x_2 & \\cdots & x_{16384}\\end{bmatrix}^\\mathsf{T}$.\n", + "\n", + "If you are not comfortable using math symbols for linear algebra, it's always nice to expand the symbols with real vectors and matrices. Let's do it with the model we'll use later in this lesson:\n", + "\n", + "$$\n", + "\\begin{aligned}\n", + " \\left[\\begin{smallmatrix}\n", + " z_1^1 \\\\ \\vdots \\\\ z_{64}^1\n", + " \\end{smallmatrix}\\right]\n", + " &=\n", + " \\operatorname{logistic}\n", + " \\left(\n", + " \\left[\\begin{smallmatrix}\n", + " W_{1,1}^0 & \\cdots & W_{1,64}^0 \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " W_{16384,1}^0 & \\cdots & W_{16384,64}^0 \\\\\n", + " \\end{smallmatrix}\\right]^\\mathsf{T}\n", + " \\left[\\begin{smallmatrix}\n", + " x_1 \\\\ \\vdots \\\\ x_{16384}\n", + " \\end{smallmatrix}\\right]\n", + " +\n", + " \\left[\\begin{smallmatrix}\n", + " b_1^0 \\\\ \\vdots \\\\ b_{64}^0\n", + " \\end{smallmatrix}\\right]\n", + " \\right)\n", + " \\\\\n", + " \\hat{y}\n", + " &=\n", + " \\operatorname{logistic}\n", + " \\left(\n", + " \\left[\\begin{smallmatrix}\n", + " W_{1}^1 \\\\\n", + " \\vdots \\\\\n", + " W_{64}^1 \\\\\n", + " \\end{smallmatrix}\\right]^\\mathsf{T}\n", + " \\left[\\begin{smallmatrix}\n", + " z_1^1 \\\\ \\vdots \\\\ z_{64}^1\n", + " \\end{smallmatrix}\\right]\n", + " +\n", + " b^1\n", + " \\right)\n", + "\\end{aligned}\n", + "$$\n", + "\n", + "Don't forget the transpose symbol in the equations." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "If we are computing the predictions of $N$ images at once, then we have\n", + "\n", + "$$\n", + "\\begin{aligned}\n", + " \\left[\\begin{smallmatrix}\n", + " z_1^{1, (1)} & \\cdots & z_{64}^{1, (1)} \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " z_1^{1, (N)} & \\cdots & z_{64}^{1, (N)} \\\\ \n", + " \\end{smallmatrix}\\right]\n", + " &=\n", + " \\operatorname{logistic}\n", + " \\left(\n", + " \\left[\\begin{smallmatrix}\n", + " x_1^{(1)} & \\cdots & x_{16384}^{(1)} \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " x_1^{(N)} & \\cdots & x_{16384}^{(N)} \\\\\n", + " \\end{smallmatrix}\\right]\n", + " \\left[\\begin{smallmatrix}\n", + " W_{1,1}^0 & \\cdots & W_{1,64}^0 \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " W_{16384,1}^0 & \\cdots & W_{16384,64}^0 \\\\\n", + " \\end{smallmatrix}\\right]\n", + " +\n", + " \\left[\\begin{smallmatrix}\n", + " b_1^0 & \\cdots & b_{64}^0 \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " b_1^0 & \\cdots & b_{64}^0 \\\\\n", + " \\end{smallmatrix}\\right]\n", + " \\right)\n", + " \\\\\n", + " \\left[\\begin{smallmatrix}\n", + " \\hat{y}^{(1)} \\\\ \\vdots \\\\ \\hat{y}^{(N)}\n", + " \\end{smallmatrix}\\right]\n", + " &=\n", + " \\operatorname{logistic}\n", + " \\left(\n", + " \\left[\\begin{smallmatrix}\n", + " z_1^{1, (1)} & \\cdots & z_{64}^{1, (1)} \\\\\n", + " \\vdots & \\ddots & \\vdots \\\\\n", + " z_1^{1, (N)} & \\cdots & z_{64}^{1, (N)} \\\\ \n", + " \\end{smallmatrix}\\right]\n", + " \\left[\\begin{smallmatrix}\n", + " W_{1}^1 \\\\\n", + " \\vdots \\\\\n", + " W_{64}^1 \\\\\n", + " \\end{smallmatrix}\\right]\n", + " +\n", + " \\left[\\begin{smallmatrix}\n", + " b^1 \\\\\n", + " \\vdots \\\\\n", + " b^1\n", + " \\end{smallmatrix}\\right]\n", + " \\right)\n", + "\\end{aligned}\n", + "$$\n", + "\n", + "You may notice that we use $W^\\mathsf{T}\\mathbf{x}$ when we have a single image and use $XW$ when we have multiple images ($N$ images in this case). This is because we use a column vector to describe a single image, but we use a matrix to represent multiple images, in which each row vector represents one image." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "If you are confused about these linear algebra calculations, writing down the shapes of these vectors and matrices may be helpful." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Now, let's write the code for our neural network model:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "def neural_network_model(x, params):\n", + " \"\"\"A fully-connected neural network with L=1.\n", + " \n", + " Arguments\n", + " ---------\n", + " x : numpy.ndarray\n", + " The input of the model. It's shape should be (n_images, n_total_pixels).\n", + " params : a tuple/list of four elements\n", + " - The first element is W0, a 2D array with shape (n_total_pixels, n_z1).\n", + " - The second elenment is b0, an 1D array with length n_z1.\n", + " - The third element is W1, an 1D array with length n_z1.\n", + " - The fourth element is b1, a scalar.\n", + "\n", + " Returns\n", + " -------\n", + " yhat : numpy.ndarray\n", + " The predicted values obtained from the model. It's an 1D array with\n", + " length n_images.\n", + " \"\"\"\n", + " z1 = logistic(numpy.dot(x, params[0])+params[1])\n", + " yhat = logistic(numpy.dot(z1, params[2])+params[3])\n", + " return yhat" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The loss function is the same as in lesson 5, except we now consider both $W^0$ and $W^1$ in the regularization: " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "def model_loss(x, true_labels, params, alpha=1.0):\n", + " \"\"\"Calculate the predictions and the loss w.r.t. the true values.\n", + " \n", + " Arguments\n", + " ---------\n", + " x : numpy.ndarray\n", + " The input of the model. The shape should be (n_images, n_total_pixels).\n", + " true_labels : numpy.ndarray\n", + " The true labels of the input images. Should be 1D and have length of\n", + " n_images.\n", + " params : a tuple/list of two elements\n", + " The first element is a 2D array with shape (n_total_pixels, 1). The second\n", + " elenment is a scalar.\n", + " alpha : float\n", + " The weight of the regularization term. Default: 1.0\n", + " \n", + " Returns\n", + " -------\n", + " loss : a scalar\n", + " The summed loss.\n", + " \"\"\"\n", + " pred = neural_network_model(x, params)\n", + " \n", + " # major loss\n", + " loss = - (\n", + " numpy.dot(true_labels, numpy.log(pred+1e-15)) +\n", + " numpy.dot(1.-true_labels, numpy.log(1.-pred+1e-15))\n", + " )\n", + "\n", + " # regularization\n", + " reg = (params[0]**2).sum() + (params[2]**2).sum()\n", + " \n", + " return loss + alpha * reg" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Initialization" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "As usual, we rely on the `grad` from `autograd` to get a function that calculates the derivatives:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# a function to get the gradients of a logistic model\n", + "gradients = grad(model_loss, argnum=2)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "For our convenience, we use variables `nz0` to store the length of input vector $\\mathbf{x}$ (recall that $\\mathbf{z}^0\\equiv\\mathbf{x}$) and use `nz1` to represent $n_{z^1}$:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# number of elements in z0 (i.e., x), z1, ...\n", + "nz0 = images_train.shape[1] # i.e., res * res\n", + "nz1 = 64" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Our regular weight initialization is not working!!!" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Intuitively, we should initialize the parameters with zeros, just like what we did in previous lessons. However, this is not going to work for neural networks when $L \\gt 0$. We will show you why." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# initialize parameters\n", + "W0 = numpy.zeros((nz0, nz1))\n", + "b0 = numpy.zeros(nz1)\n", + "W1 = numpy.zeros(nz1)\n", + "b1 = 0.0" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "When we have all-zero elements in parameter matrices, the gradients of $W^i$ for $i \\ne L$ will be zeros. In our model ($L=1$), that says the gradient of $W^0$ will be zero. Without dragging you into messy mathematical equations, we can show it by using the `gradients` function to calculate the gradients:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", + "print(\"Gradients of W0 are zeros:\", numpy.allclose(grads[0], 0.))\n", + "print(\"Gradients of W1 are zeros:\", numpy.allclose(grads[2], 0.))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "And how does this issue matter? Recall how we perform gradient descent -- we update parameters by substracting gradients from the current values of parameters:\n", + "\n", + "$$\n", + "\\text{new }W^0 = \\text{current }W^0 - \\text{learning rate}\\times\\text{gradients of }W^0\n", + "$$\n", + "\n", + "So if the gradients of $W^0$ are zeros, then gradient descent will not change the values of $W^0$ at all. In other words, our model will not improve no matter how many iterations we run for optimization." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Xavier/Glorot initialization" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "To resolve the issue, we need non-zero initial values for parameters. Xavier initialization (or sometimes called Glorot initialization) is the most common method to initialize parameters. The mathematical proof of why it works under the hood requires knowledge in statistics and probabilities. You can check reference [1] if you're interested, but here we focus on how to do it instead:" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "$$\n", + "W^i = \\text{random numbers uniformly drawn from} \\\\\n", + "\\text{interval }\\left[-\\sqrt{\\frac{6}{n_{z^i}+n_{z^{i+1}}}}, \\sqrt{\\frac{6}{n_{z^i}+n_{z^{i+1}}}}\\right]\n", + "$$\n", + "and\n", + "$$\n", + "\\mathbf{b}^i = \\mathbf{0}\n", + "$$\n", + "where the bold-face zero, $\\mathbf{0}$, denotes a vector of zeros." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# initialize parameters\n", + "scale = numpy.sqrt(6/(nz0+nz1))\n", + "W0 = numpy.random.uniform(-scale, scale, (nz0, nz1))\n", + "b0 = numpy.zeros(nz1)\n", + "\n", + "scale = numpy.sqrt(6/(nz1+1))\n", + "W1 = numpy.random.uniform(-scale, scale, nz1)\n", + "b1 = 0." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Now if you check the gradients of $W^0$ using the function `gradients`, they are not zeros anymore." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Finally, let's take a look at the initial performance of our model:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# initial accuracy\n", + "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", + "perf = performance(pred_labels_test, labels_test)\n", + "print(\"Initial precision: {:.1f}%\".format(perf[0]*100))\n", + "print(\"Initial recall: {:.1f}%\".format(perf[1]*100))\n", + "print(\"Initial F-score: {:.1f}%\".format(perf[2]*100))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Optimization" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The optimization is the same as how we did it in lesson 5, except that now we have to update `W0`, `b0`, `W1`, and `b1` (in lesson 5, we only updated `W` and `b`):" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [ + "# learning rate\n", + "lr = 5e-4\n", + "\n", + "# a variable for the change in validation loss\n", + "change = numpy.inf\n", + "\n", + "# a counter for optimization iterations\n", + "i = 0\n", + "\n", + "# a variable to store the validation loss from the previous iteration\n", + "old_val_loss = 1e-15\n", + "\n", + "# keep running if:\n", + "# 1. we still see significant changes in validation loss\n", + "# 2. iteration counter < 10000\n", + "while change >= 1e-5 and i < 10000:\n", + " \n", + " # calculate gradients and use gradient descents\n", + " grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", + " W0 -= (grads[0] * lr)\n", + " b0 -= (grads[1] * lr)\n", + " W1 -= (grads[2] * lr)\n", + " b1 -= (grads[3] * lr)\n", + " \n", + " # validation loss\n", + " val_loss = model_loss(images_val, labels_val, (W0, b0, W1, b1))\n", + " \n", + " # calculate f-scores against the validation dataset\n", + " pred_labels_val = classify(images_val, (W0, b0, W1, b1), neural_network_model)\n", + " score = performance(pred_labels_val, labels_val)\n", + "\n", + " # calculate the chage in validation loss\n", + " change = numpy.abs((val_loss-old_val_loss)/old_val_loss)\n", + "\n", + " # update the counter and old_val_loss\n", + " i += 1\n", + " old_val_loss = val_loss\n", + " \n", + " # print the progress every 10 steps\n", + " if i % 10 == 0:\n", + " print(\"{}...\".format(i), end=\"\")\n", + "\n", + "print(\"\")\n", + "print(\"\")\n", + "print(\"Upon optimization stopped:\")\n", + "print(\" Iterations:\", i)\n", + "print(\" Validation loss:\", val_loss)\n", + "print(\" Validation precision:\", score[0])\n", + "print(\" Validation recall:\", score[1])\n", + "print(\" Validation F-score:\", score[2])\n", + "print(\" Change in validation loss:\", change)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Finally, let's check the final model performance:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# final accuracy\n", + "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", + "perf = performance(pred_labels_test, labels_test)\n", + "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", + "print(\"Final recall: {:.1f}%\".format(perf[1]*100))\n", + "print(\"Final F-score: {:.1f}%\".format(perf[2]*100))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Awesome! Compared to lesson 5, simply replacing the logistic regression model with a fully connected neural network model with $L=1$ improves the F-score from $86.4\\%$ to $91.6\\%$. The precision is now $92.8\\%$, meaning whenever our model predicts a casting part being defective, $92.8\\%$ chance it's correct. And the recall is now $90.4\\%$, which says that our model misses around $10\\%$ of defective parts." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Of course, the performance may still be unsatisfying. However, one benefit of using neural networks is that we can control the complexity of our model. For example, we may get better performance if we increase $L$ and the lengths of intermediate results $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. \n", + "\n", + "What we didn't address in this lesson, however, is that increasing the complexity increases the difficulty of optimization as well. Optimization may need much more time to converge or may not even find a satisfying solution. So it's not like we can freely increase $L$ and the lengths of intermediate results to whatever values we like. One way to ease the optimization when having high $L$ and lengths is to use advanced optimization methods. In lesson 8, we will see more different optimization methods. Another approach is to use other types of neural networks, such as a convolutional neural network, which is a more popular choice for computer vision applications. We will talk about this in a later lesson. " + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Conclusion\n", + "\n", + "In this lesson, we learned:\n", + "\n", + "1. the concept of fully connected neural networks,\n", + "2. how to code a fully connected neural network, and\n", + "3. how to initialize parameters of a neural network.\n", + "\n", + "We hope this lesson gave you some sense of what neural networks and deep learning are and how they are the same as or different from linear/logistic regressions." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Reference\n", + "\n", + "1. Glorot, X., & Bengio, Y. (2010). Understanding the difficulty of training deep feedforward neural networks. In Y. W. Teh & M. Titterington (Eds.), Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics (Vol. 9, pp. 249–256). PMLR. http://proceedings.mlr.press/v9/glorot10a.html" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# Execute this cell to load the notebook's style sheet, then ignore it\n", + "from IPython.core.display import HTML\n", + "css_file = '../style/custom.css'\n", + "HTML(open(css_file, \"r\").read())" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.5" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/scripts/lesson_5_functions.py b/scripts/lesson_5_functions.py new file mode 100644 index 0000000..33611e0 --- /dev/null +++ b/scripts/lesson_5_functions.py @@ -0,0 +1,110 @@ +#! /usr/bin/env python3 +# -*- coding: utf-8 -*- +# vim:fenc=utf-8 +"""Functions from lesson 5. + +Usage: + from lesson_5_functions import * + +This will make the following functions available: + - logistic + - classify + - performance +""" +from autograd import numpy + + +# doing `from lesson_5_functions import *` will import these objects +__all__ = ["logistic", "classify", "performance"] + + +def logistic(x): + """Logistic/sigmoid function. + + Arguments + --------- + x : numpy.ndarray + The input to the logistic function. + + Returns + ------- + numpy.ndarray + The output. + + Notes + ----- + The function does not restrict the shape of the input array. The output + has the same shape as the input. + """ + return 1. / (1. + numpy.exp(-x)) + + +def classify(x, params, model): + """Use a logistic model to label data with 0 or/and 1. + + Arguments + --------- + x : numpy.ndarray + The input of the model. The shape should be (n_images, n_total_pixels). + params : a tuple/list of two elements + The first element is a 2D array with shape (n_total_pixels, 1). The + second elenment is a scalar. + model : a callable object + The model that takes in `x` and `params` and then returns the probabilities. + + Returns + ------- + labels : numpy.ndarray + The shape of the label is the same with `probability`. + + Notes + ----- + This function only works with multiple images, i.e., x has a shape of + (n_images, n_total_pixels). + """ + probabilities = model(x, params) + labels = (probabilities >= 0.5).astype(float) + return labels + + +def performance(predictions, answers, beta=1.0): + """Calculate precision, recall, and F-score. + + Arguments + --------- + predictions : numpy.ndarray of integers + The predicted labels. + answers : numpy.ndarray of integers + The true labels. + beta : float + A coefficient representing the weight of recall. + + Returns + ------- + precision, recall, score : float + Precision, recall, and F-score, respectively. + """ + true_idx = (answers == 1) # the location where the answers are 1 + false_idx = (answers == 0) # the location where the answers are 0 + + # true positive: answers are 1 and predictions are also 1 + n_tp = numpy.count_nonzero(predictions[true_idx] == 1) + + # false positive: answers are 0 but predictions are 1 + n_fp = numpy.count_nonzero(predictions[false_idx] == 1) + + # true negative: answers are 0 and predictions are also 0 + n_tn = numpy.count_nonzero(predictions[false_idx] == 0) + + # false negative: answers are 1 but predictions are 0 + n_fn = numpy.count_nonzero(predictions[true_idx] == 0) + + # precision, recall, and f-score + precision = n_tp / (n_tp + n_fp) + recall = n_tp / (n_tp + n_fn) + score = ( + (1.0 + beta**2) * precision * recall / + (beta**2 * precision + recall) + ) + + return precision, recall, score diff --git a/scripts/load_casting_data.py b/scripts/load_casting_data.py new file mode 100644 index 0000000..1039e67 --- /dev/null +++ b/scripts/load_casting_data.py @@ -0,0 +1,84 @@ +#! /usr/bin/env python3 +# -*- coding: utf-8 -*- +# vim:fenc=utf-8 + +"""Load metal-casting datasets. + +Usage: + frpm load_casting_data import * + +Then the following variables will be available: + - res: image resolution in both x and y + - n_ok_total: total number of images for normal parts + - n_ok_train: number of images for normal parts in the training dataset + - n_ok_val: number of images for normal parts in the validation dataset + - n_ok_test: number of images for normal parts in the test dataset + - n_def_total: total number of images for defective parts + - n_def_train: number of images for defective parts in the training dataset + - n_def_val: number of images for defective parts in the validation dataset + - n_def_test: number of images for defective parts in the test dataset + - images_train: training dataset, 2D array with shape (n_ok_train+n_def_train, res*res) + - images_val: validation dataset, 2D array with shape (n_ok_val+n_def_val, res*res) + - images_test: test dataset, 2D array with shape (n_ok_test+n_def_test, res*res) + - labels_train: labels for the training dataset, 1D array with length n_ok_train+n_def_train + - labels_val: labels for the validation dataset, 1D array with length n_ok_train+n_def_train + - labels_test: labels for the test dataset, 1D array with length n_ok_train+n_def_train + - mu: the mean values of training data + - sigma: the standard deviations of training data +""" +import numpy +import pathlib + +# define what will be imported when doing `from load_casting_data import *` +__all__ = [ + "res", "n_ok_total", "n_ok_train", "n_ok_val", "n_ok_test", "n_def_total", "n_def_train", + "n_def_val", "n_def_test", "images_train", "images_val", "images_test", "labels_train", + "labels_val", "labels_test", "mu", "sigma" +] + +# path to the repository folder +root = pathlib.Path(__file__).parents[1] + +# read in images and labels +with numpy.load(root.joinpath("data", "casting_images.npz")) as data: + ok_images = data["ok_images"] + def_images = data["def_images"] + +# get the number of images and image resolution +n_ok_total = ok_images.shape[0] +n_def_total = def_images.shape[0] +res = int(numpy.sqrt(def_images.shape[1])) + +# numbers of images for validation, test, and training data +n_ok_val = int(n_ok_total * 0.2) +n_def_val = int(n_def_total * 0.2) +n_ok_test = int(n_ok_total * 0.2) +n_def_test = int(n_def_total * 0.2) +n_ok_train = n_ok_total - n_ok_val - n_ok_test +n_def_train = n_def_total - n_def_val - n_def_test + +# split dataset +ok_images = numpy.split(ok_images, [n_ok_val, n_ok_val+n_ok_test], 0) +def_images = numpy.split(def_images, [n_def_val, n_def_val+n_def_test], 0) + +# combine images of defective and normal parts +images_val = numpy.concatenate([ok_images[0], def_images[0]], 0) +images_test = numpy.concatenate([ok_images[1], def_images[1]], 0) +images_train = numpy.concatenate([ok_images[2], def_images[2]], 0) + +# calculate mu and sigma and the normalize datasets +mu = numpy.mean(images_train, axis=0) +sigma = numpy.std(images_train, axis=0) + +# normalize the training, validation, and test datasets using mu and sigma +images_train = (images_train - mu) / sigma +images_val = (images_val - mu) / sigma +images_test = (images_test - mu) / sigma + +# labels for training, validation, and test data +labels_train = numpy.zeros(n_ok_train+n_def_train) +labels_train[n_ok_train:] = 1. +labels_val = numpy.zeros(n_ok_val+n_def_val) +labels_val[n_ok_val:] = 1. +labels_test = numpy.zeros(n_ok_test+n_def_test) +labels_test[n_ok_test:] = 1. From 1f9fa8e480dd1f80924f162d977fd6e2d39ba82b Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Wed, 4 Aug 2021 02:18:18 -0400 Subject: [PATCH 2/8] added an appendix to explain neural network terms --- images/neural-networks.png | Bin 0 -> 68969 bytes images/neural-networks.svg | 10131 ++++++++++++++++++++++++++++++++++ notebooks_en/Appendix.ipynb | 155 + 3 files changed, 10286 insertions(+) create mode 100644 images/neural-networks.png create mode 100644 images/neural-networks.svg create mode 100644 notebooks_en/Appendix.ipynb diff --git a/images/neural-networks.png b/images/neural-networks.png new file mode 100644 index 0000000000000000000000000000000000000000..ef97432c6418c127833208110d670ed79280196f GIT binary patch literal 68969 zcmd4(_di_U_x}%L5MA`qLy+jb_ZU6OXkqj|h~9gP=tPw0T}YG=hUg=r_h7VWQKR=> zuJe4ozn}l%`T;j*&RJ)ly;r$E);`2)Ybq1s(cqz>p%JR8Ky=X1FyLrt=vMo*#8FVPVg3s^5MmFOqZUmvsG2`Z8V{26vcb@eB>k6}9;|n%yR< z?$KE=7W8rW&*joU&h3Hp<)3;mT~gP8*E4J>BWHO!B4v$wSvUM~te`(z^(1mHYPoAG zetL-05%m>XsjrA`TLkK++q#R$76?^vNq!XpQCK!@n{ z;ZMTSk^iR2M*hDu8s-=@bnrFhY1I<27_fsZT=m33e$*N>4a7%7J29Wd{%_<8SqYDG zYAatz=NUmk;u7w36n|f?gsiAoC>D%uf?Cma!*J6nhE~CUOu{mpu7kaMfYEy+ySM$0 z4RM`m(THP|Ia0dA>;LbOw*}~C#eql0-6zIcf;9EeC8>PT`ydB`G1fgzgvTc8B(ZM< zCC!(i8@4m{d1I(6Wb~}MP02hDC%G4Mp>FZcN~RBkb{*Eoh8e8lLcS4nSV z?Q->3+GO$OtK`y=K~;oONjy*c$l+UV@07~8;P?|PZfL3&Rw#Q@WM`zvTw9XFjx#}3 zEL&nAnIT$((u=Pp7*%syhg)H{VIl#NP9Y%fi2RWf#nE<&TJNJ_H zsccnXMo*qDhVp9V+-vVT_{a{hknIiKR743cNZcBr4zn&Varb3`^+{hB@glfFn`P6= zaWa|?@PQ@y$d=fPvfU!y<@2va(iQmrUi-FL!I)IM6}W$!$3wg{Rvy`wzm`e#HJ6dw zu~AW@Co5EpFbJWQLqEYiZj02UNFOB;eo?r5OXe@Lj5DH|davCUi&FChJUKaf+ekjx$`C;X0fU$}JMj&BTW(<}0mnB(9h+qj7-luo$crXe)K3aPVBvreo6uK2Z?5wgz-j z7E0og)TM0aTY`Un*n;g%2><$7dpmbL_9n`ON%6&s|5WKWlrB+Du_&DIIPoZx^0R1(=I|%ZG_*xIG?v)Xqa6l2Gshy+ARE-6o}8j@{p`s^q&?7=(lg1IHo&!XLtWyo(~gg5?4 z^x9GD3sqln2v{3GHxZ*q11^&_DxXafhaFz}kD>Y2?qs^_*(%a{yj9nFR4I`$$$D_7 zM7WwtM(E!TH%K|V63NK62o#@xrey0_tqAQpFgoOxE0LD0WQMvE6*jdY`sd-|A3HMY zg#yf5$0KS*3!x3c&oenjScI|#B{7`PCBR%?;~@82o@eCdHTu0p@?yc<_KWc4ZHxm& zD(&PtNA@&6ajGkaquXdCTiRouYyxT$bneXMs~+Sm6epJYYSm=raZpP6OHlayWu+4(Yr(Z$k>XLR7-JbvRN)6uY37`H*OW! zp;ps>yI&MPK8^uEc~P%iwCkg$v7Wsu;%0YM6iz#{)6 zrNYX6>NIoq^=PDG*P*OJ)33oIl99S5T1RfFry9GYSZmODQ*KXzmg>}Rp~H~dKCa?J z#$f7br77EF5F^?`k-<1YPI1qeYgtE?BW8=YGTzGM5?xrVduy>&r7br%`my#6&WkrB=TQ2}6=#ypj{w8z3*qQ@(%~vF(3?V zLYH|bbZL2WAV=4_6saz5C#DUdZFnXzFnY(Hj|UYpvah&zz7?iUicA?i{`jCzb-K9@be!??kRcg{O>px)UZ5Xz3QAFluh7Q?Qr4D^vTg_jyR z$?7^Vyqv1tAP!s%4`2Kl-?<*zCAd$E~O)ExKt-ft^A#XHFxT z3;XFE=K6l~?|k*`sF~oO6S5-8)izFg#vh)FCokPQu($BlA$cobzR{l}LL_(>KD-Ea zkZUse?)t}by|GGcCmGzbG4IP7NViw!FK>hB_4|XLDIWZ6q)g=buaZO`*6oPZM&uv=d@^q& ztgpMNPt3!Zv{T0bqi0%(Fd$P4?r6WQ^~cq($SCH_0kge+5d3m=IRE&i~UDe0D*=5tMH)I_i+5*F93WpDi!`~>s`OTlh>19XtE;x5AlHW z2rPR`Sdmj){vK4Y1Z{7!52gg=%SEbQRQ#|}te7U1`@b8| zfy#=;mg$W;3ldSw?KIVD|8e1eA61}@AW5eJE_nDT3>mK|9K~Dn^wY)u-#C8lMH-Z< zPL@;!SdDP-DNvc#HRIqJy8tQ1B{r#vkn3H-o>bXa$(`CIqmUP)pY^ki!O2X7SmvY&^W z7h7wKO0X@`<~l+8ze^U|3$HjKJI4>}@%^xWfTvj-UwOw{C7?GdcxRY{Jhe^j0G@Sw zIZpn4FXg+81eB-vlBHQQW(JH;hy91ptB~>sXLu}=SvLvc8kF@dzXZ-wvd;rTp{+N2 z8<}lH7ldO%^cv@whA6t^WN!GBX73_z1Qf?6N%_xv^%=lXTO}piY={I zIHovI=zpR(pib}n65T5-U)GBmzT*hljq*zzpL?qRF$3g^zBO*mknj-^J~<)%Hyx=_ z@fcpxn2BaWSS2?yxU86dT|xaM7mG`dadvkidKQor#3RVWZLln(SgIS<{WTCq3_Ef4 zbXVL*U+_$3pbWSxc7SJvxso0f%QCKoAA(##jG)e`&u<1){}lyqvuA;)PF-aFYl8@{PxI= z7$NGbDlM&Rnrcn6{FoPc&=>y4)jJ$)e2LsBmfHq{>8t9Ym`S#_}fc z(u<~t_6;V7GeT+cCiqs~>RAR6&Q`bl1#Tb}L_drV)@)QWs=`AdM?TNF3uv9iuFq?3R>B{qfiM;`0DMpfkduN{fm z1-nts6VD?2D>i*imdB~NUqnNifCJNI|#QH)jYB3~h#`c_Fg@n|WCD@@w1hWRfar7XEj z=g9T0;dO5@b=dM}ne<_ptwO`8v{vF)1p7}IQWsA1A@P6qY~V`O2UwD<7+=}kcFWf+b=#9s z$jDAbvAE$)Yd3T#L7}O7^Qp`CImbE?M^^u=8v_xisN;&Cw_IiV3}3xbOurtMVo-TB zZz-dO3y&~r2*%~W*iuP1u7sbt?y2LC>!*}|>`uB01&cHb8nO=1rKkV)OdyXOKn92K1GTH>0VBCgZL;BN{ zzgEGeCeaxKUwVY3K(_c^Z`(@eBc}L8L5KWe>S8XDO2uZg%Vc}eo7CbwT(OQHK;u8l z%_|$9dG~Ao@Qucxyz)R`xr<>_D;KWyQGMmcaoiZRXLl`pY~?ane(W1NKfNj%o|y=| zyyh4pVSP~~fs)x^_(8docG*|U`c+O$E&}YwKWcTjj=^mW^AUPs z#uFh18M5<-<0^e!D-A{otJ@5h%U`)w=-tV7tf8EmIlE|*XwiiWZ0g!WGj&Y4fh;uI zC!d~<0&Sp`Zxzc%)z`kGa(#sHs6~QEet1BPlY+#u_R1lgx&!u-5G?Fe_yB*C7Jw9* z+7n4@^uADH90XMY)@wN(17xb^BW&F9cHI1>=Uod}q_7ArmREvqqr7B*&d~|=nUO@{ ze8l=nYsFhmT+!Xzl0h2i21H7okEODBBM#;3_SL{L_^%?4CkgnqRL!_ zS5oCb6Y7c?_NkVV(_85G8jU*?b8DG$(f{QFQxMLvuoqDcgf*KbTc#>~zCT)mem@XP z9WE2$c%zUmTZH>2FP%j+8Fcx^IEFOll|1M?&p*7#-qc6z=Ze2qc`@Z&EW=}cJ9A5~ z?8F~yS@tr~NO>WXV8^g#fTV(&~juk7)iTfz0PiC0&fbHE;$6~d1bN4rHh z0eX1zBHJ8(habjv-3WYSB4s2I*}A8<+nE#cUIRnE_%(Ibkgad_M6b4%MwmNBAq`xs zrrVcoY%Pc@=Ir+3Y@47}Ps)f#!ktLrsVvp~8x_4&r*TaaVn|tG_znDlDdJRuE;38D zgz^fu!Q?gka_5PVoY_y-&e5JcNM~RZNw&y_N5E~iD0)+oOoTOaMiH1B`R%4doM2?y z!=d_mB~fo?vb!wjqt zx?@B~^rj y^PgEw=?0%?Owx5&mI@+F!<1j$+Hvwd)aQM1G+XV%XaTR{V}43}Y-d z_#%tmch>9&%7$K!uAJ&t9ISt$eyv$iOn~UMtnhT@7Sqjmm4*ki>6UR0^ zLphp_m~J1-=8zmCL?C^?VHJhioVf`c{m`vxx;L+K9B+%A6Iw9N)D%1^j$UI@-Ad4~ zx(mQ^_T;8?F2UJ_&IB!GR|->)PEV#VrIL3j`8%I{`XQ1-8?$uA$Oh$rqY?ilospm` z^N>W)o17_lwHb!9eNM`L;U4Q$?|$m+YB%}9-5DPSmVTY2m>PXC9XAGHd|w&f`cFFc zAuB)qwJ`np7zF=ID_paTkULt37gAlAmzZ5-DpcML1xLzoBK8=LdA-^)^{s9f!c;*& zmK5QvSkVWC!DvAxK%#p3sd%9-%*EJ9XpoId1lx?YNmY{U%Xs2vu6{mJ;eA{325yD4 za+5{^o7u)Vfhjz)iNLx*1QvT_1x3Stm+_iBZDd|89yw#`%_|A(Hr z{H}90q7}MGEYXab!G{W-h7f|0$B#-QHI~PmE3X=^-kJ+2mRloT>zB-vVB<>=g?w{7 zXugL(_4JPKS71ecdI|X#4n-_>he8W*XJqM;e(vD}LH3ox4k5{sBUPmH1)x6>92EB( z>;T%7EsO^q$vBSV=6H{vgz}b5Ba16bVZ(l=4kC&}u3ma8Fxo=SgX?vz>+jM> zCRjG?5YZjue>Z?S%G8*s6QW-geqR*dA0$+a2jvlvN9TOmfs+?L3LUXkZQ|h6ZXyv4 zfj?&UA<36asuD9;?pWQVU|p+BQ&!oo_O_s@&L1R;5f8fk0_lS~LMaF9;Ca=b!=V{U zt|}c*tQ?Q(vHV&SPKlJ8lshr+=$=kBdrm&Y8l}y+5kGymEbB3WWAdJdoB_f(i|^TR zOOW0D8JtBY3cY#S^04Yg{^IMgHP}st#tRjWN;VT!ZqO=byg->2R-!JJ;~imZA#CNI+nu>9bT({Q%5#1&Ns z7aE_2%jB&^oDSpaBDqH8--s*ZZ9&U>UN*mHSnl3?b6^#rSMp zD1b7NMTYN6YB_R_aa5U#8h|0$YPd@|p99*%=-0KLp)<`gJ_EWT?~Kld!2!hl{Tw?k7;6SoUnF}W|@ zpWUpu#B$gvm!w!U8;H2;IhaSl+(MIG=8_eIx0MRzUnZA;oWVc-H>GjMYkGR_0py6F zL?u(k&rc5X;$?EEitrmA&&5M1rSaR#y2eV_b6KfbHuKRq>YOQfE+S9cD0u3M6Rp#H&i~BY1m9rS0-=7W(MY$7U(%3(@Vo(u2VjT7e z;eU?xI2tT;6pwP%Ma_nU=k!Nm1(25tp<;1^R~zDV=dWKJZ9E+lCD8vtG<-bQvTf+W zDuEj+_-pt52zy1D-OI_a(beepK}&C-j12^2+jU|2*;|>f8!^-^gW=p$^yRd(LCBAr zTd!9}l!T1-lu689=4d^UHLl2=uIUWI$!tCB2rT%bTmGIx4(iK78q1HGe|g)-J8sj{ zY?#FH>ayhFJ>pk1@+ilp#o_+H?I}A7L@e2h5N&5_x`{fAu83@I)ywUYWD$DLVgAH- zlATrPcM1BSVYzCerb2Nc-eRrqV>D|D7O%NOL!qZR%XfC?Lp-enKWAhwAkP&9>0)VM zEl}B?CVMI8w^cF)^zqm-no1yu3`H*QvVj9Q2q(b-(iDm(uKZpRRBL>6ONKRn&v5=6 zMvjnEG$g8B3fPr~JI|uCoy+_BigrCR*gpCRjolK$&989I87qV{K(F&85M}lC^%;`= zAM()NW8~%KeQ_N6kPJMtT~OFhI@zV?%KZPrRx1ilr1az~Z8JdKZ>kJa9N{>V!i^zf zXX6Ui;SrPtvy6c&Y(rt`E4V_boDEOObq zb(|8gGew-|Th=cf+{Bn!Q>DW602XgDQSq@#ArO-E$$fh*IJw*qI& z&+{U)zd40Zho4c~=Q2%wj?iZRxyM*m%jQPa5_KGpXVzq5p!_C8Q`}R`qh7#IQ97h5 zDv3H&J~agesw^2C&}P%^Vj6sKH#`zdyht&5`$|hc+9Q&x#L9c68%)z9z*d@iprFM( z|GDaNM&UW%X%>~dr-?K6AM%e}AX@_Ie>Cp}EH9F+g7D6jTf-mN?Gr0{s!11U)6W*Y z#CFmv`jR!qRnPHm%5!_h+w#&!RX2Ny^|NKz<{wgy0}B)$+A0HW?qxV%84X$z96_IR zoV{nt^HM!J4GuU-OO&S<4UJ&?*3KZ7xs<(Q3D*yMP52dNA%$S<*Tw~&GHMp{2QByM zc`|7-H4)1M!MJ2GI#v$GW5O$>%>RY|;ITEH+mmbqrUG6knchOBaKUiXaIzQ~CSMVk zJMH3pbAj070`76Pwa^Qh%9{fH_+zAGVnxAN=?6^-d<~uemK&`xxoXEw$mz$hgKD#P zFm;`Zu{m1!x*~TJ{Tt@U5b39OkNtfHz(Ibv)gKcAg(p+D^-DQ8CQ%nTTjO+!TS_nN zcugMN#Nx=Nuj)W8v$7xWjFy#)yOh-Hmn88K-+2|P&$GM;Key3i0HyN9PgQSq~P^moEa~d$bs9|xFNx27MZq#eQ+V_9UDuaS5N_U}g5+&0C`nmFqpsUQ1zz0@*={l3T?kWwFX(L5k9hGi zlz@9bZP-yTJO9bm66vTBJC^h1ibM`1h(5YxNwsinkcTPPNib8`^zQ$?3oxa%&D0Qb z$d0@To(BsF-ajGd`IuisgEt@kPFKU8j}NQ;agtdg6eG`C2BI5mJIknnl~H_mlLO_1 zM;9xlXeHP^Dpq#RkeSQ5{{DkA8q5Q;A1WHg;=w%*>JEDmzA&y&0B_vnIg<>=Uwmc4 zaCeWL{k;qf%npxS(@UTm=e{X;6!Fg*T2LLR-MKkdNcL8=MbU@C&3U}H*toDnUCdRO zykRJ2RJeSJZ`+U2r|azPUwBfR&kB~8Oe>cD5@pfyW1ej;uOrsJ+6|`gUBrB>*~?$@ z+mH~puX?aE^maN#LlBm!En!+HdMLrFRAdxq|6$uQV@Sd#*G{X|HM6uDZ1bF>Oftnl zO+JuL0}JRh$DIaZ=Xb0aOvQZIAMRw3aGjI+=6|-1;7|g_LiI&9QYGy!Y7nv%$tFKm z%PAea$!VMW92kH;Mx=^eL>r^9u#(Z8Ac2qki{10^&JxXxt>TOfyQz2L*w>9TsXX)sM$vVVJ~x{$n&g=A;O0={YVYGD`o6#tuoqKhEVB68vQlyfntc z=?ytL4TZ;*!z?%)ZFZ(*31Gj=)1+^^Mzbo$bjyZlrv(VNGY+xzGfKeewE%?QbO}*7 zZM%t~Az(pr4=OE=Iu7{ritzNjBBL-t(Gnu){Jj_`#RmC8=B(iK?A8$N|sK9}n3<_9;wpFeB; zS{jkT|IP#!?leu?S?(q<=K)qeRl=X zF`}T*0hg}B4_Nc92J~z{xtID0(h*-L3}Acd$WvFwiEH|#uEC`3`g#EQ|0%FiG>ny4 z!>e0K^&_=Pmb*2Xax1CZ$6QM!t%AkrroA`EQ!9QvRW71zz8$YVP566>e&lv!1eS?X zE|i(|H;sLWQxc#0p3b7ifR_KY@u~1|J8}}($wLF&Ke?E=w0resfsx)Qb(xdHxQ+Aw_(Br%# zs}3}Z*Q|lE!?8_M?j_d&@vD$nuB{HMe*Ga`dz@xG@5?Q7TdxJOP-cZV%5Rh%X<30m zZa^7POa1D49Ph$4{byy?{eZSaG zmE%9Lu|IDAufh?Z^oVbQ6JB28AZKEm{dX^Hv`3lTNi61y2S!xgJO$!`=xQYJqTHU= zDp^~vwFd6h^;P0ia#b`1lv3D_=8BhGE{8sD(a_My6pN^(?( z@U326L7`h&a=&3x8pxd5R@I3bc@n^z486Tg6WL{vdARVF?-7=@^saifY2o!7)=@1L~TjPkYA785TCAbPP!)4mywRiS!@=eIyaUU;Q zL0Q0kcj16@E3}MYs^t^4jdEe|weL2iPv~f3?i0#0a> z?~>hYJq|=!j1(s`@n@Yc9p}2CaWt{mYhtakEmyJ*g)eSEEx`?oa$0%{EE5tg#2=p1 zxUc?1`*%yE z?07Vs4^mVWPq`dIX;`hTFn0PmzTbqz2p~#cBAIVkR4%|8tSPSZkYughK71HRW@DC&@2_NSAx(UxTYpvY&7eJ1j$-r{mp^^Z z5h(~ync@B^;XpA9_9KL&qu4jXs?sf@W283xmkPNjCs3%yy2IN?@|1=c6G7jDKM{Cu zDJx5iKA6AoU91maBErc|wX4=4#%l~W=RM6Ar;CicxVT8^8tu+uK0yc&nbKzB`=zBg z7GlrQ(Sz(uy0BSExMG$`Q--#Nvvl2{hGRDSuiO9KhKRGW_Du-7UM^x+zi5PbS3GLl z94|8(D4ePXqwP^%3~4PNfNN6eUUIvg)VpN}gp8Bf`F zbkUESelBNsKQ2g&siOUn7864tQD8j@uwhYFByQtjQI&LkMO}Qp+BOT41TI23Bitub9hu@&>t7BRYz9PafC|nb9 z28cx07o{h@;lBf$W7?RQpG3QlI(|q;fA)GNnFPqgBn_b^N>{N_-8RBP{Q`t{2~z0x zP0w90J*jOg_eqCWjdEoc`c8`GZ&FJ%>m{ULXxb>r6rj{?zoP$ zOc&&81-<~$e*Tu}AD}8Q4|_Ht5Q$Gy=JL1lDBIG}807m-oF*(H9P81Kl)dhLP;pO3 zZuu`$r98gd&&j3mtE(MG#>b`d!yxG)747Z8Hnj_>s%DbZizj>7HLvaqD@WeLy=n3P zC?=*2Nj+65D#NRhr)W~E7jcv4qBKsB$PP)A5r+eL77BTw>*SIfiG#f!` z;{W~yf|-z1fM}pCAnE3Fb+xlKiU0U%8&&Xf@-M5^+lO=JlTUPv^q6h=!XbLyPiuK3 zLV_^cqyE#}V)X0qZRy>f5|hdBs@AxR*-DdukJv`u#N zYWrE?BDkJ>pK({&{hqn%iAll@^9>yLSGE-kjHaAkf2s!LQ-kzwQtC`gPT+r(yg!@# zWE!}S(De`|Fo2H%DcFNNg3GYhVgzZ4Ld6-)pSDedLf^ZPG|PUCUN~(Vi&rt0>AOiK4nw$CCRY&Ms z@`!bIhre!9T*0YdNAslv4asMYr4C<`I1FqJUF@1a2Alv*l;b-7%~1H$-xPXm-FKQ5 z6PEs5l!3c@^+)COB!n=&}QOJMD|rZXY{meBi86hixNJ}!?rZWWasw575|PmHSV4l z#EK)}55e{~lU5E+rC7a(7#kbgs18PY$R3ghPy#Ne2H}yXKOKYI2XbZCAJGX3D}(XW zK@xBBX0~WvfHasr5L^CKRJZ2ETx;w8OXMWi$uUxtjX7%KG5tWrkfr~DfMk^ZxxC+3 zTn~+Ucr$kD@-zj1qOixcMBAf3|4)$mEf)rXKhf3O}zrA8(&! zMOwpuFQBSsPNe!oi$ZWxq2}KmDm*vm;bA|`p`l^r0#B|wI^QEu@jX_p!lI&`8M~$V zTl6*w&BXvM&I5^ZnRZ(?D_-1lGl2o1Lcq(G6$SSTa#dL7OMV!bQv%LqXK%mNgN*IX zl>Ghg%wfRhBou1m^dD_2Co|k9@Na%P_Pgsfah4jRwNLP$hP7%4WCcCnEpdRy+>lfl z1@Irau88Zo51l1b`135k%h=ypkVU=96FihFf#(m<@wqOwA)*OsFJN!p%v}7}Xf`M`J%RiY3Wdtq1BhsS%UT=E)5e( z%K*VH1CRMhL{#+g4>&dHFATsG>(bp+X13DUD5^Dc_6EED#I^>lGCd8IA3+ zG8O`+GWCB`X`gWI!q*$5H=p;C*vWbL@_sG60Bp_Uun{aL9>3K>4Jx{ja$#DiQmST7`t`TAWi1r}(cD7cP?1kptGrxsGNk4NeB1Kb^ zC52#Ae!1^szvzM8jj~M@GVsu{n0gx^i!(DzI9W*Y9u*tQUIMj_-^2W|_aoBQ650 zhyARF{h>Z~+U8=`!>$v70glWmKP108h2J<+TU#j}oPVoEh-AIa*I&`o9e>~ljr?fi zm%c9{TSWCs1vTIFVKit-(X^S#-|F!A^9c05eVW{tJa+D%F?+bqg zbbDv#hWi6wFCA5KUqSp1j7XUuAYy}mUN_h)rq;s)!1&H9YACpL9vrKK+s9`Cuq5ow z6eWxO>w-vaz^1F4) z1uDJz_53?|b@d@OLPElsUsiSQz-@lm2wX4!RX%4NJ8k6e{e#BUfXe@4-$l<6P8v{v zxG1v`@Yn3{0SfH|c=DqPOUTD^Rf2uStZQ3*p)l?@k9^yA0`K#6UYo)n{Yds11t)Np z`oGvg=G>+vGUOODrnh5%8b&PDCWndO8AA;$5B@Y)3JVLz^4k^oNj-Sfh@j@NmoFKG zdZieW`*`V)kI=Nn#zqoTQ`0T5^(a=p0~{`9nW`b;fO|?~M0VeW^hV7evZW3Xceg2- zON~j6fIa_KKubO+Gs-@!nK-K%xoYb}<+T*H6nMaX!S+ic>T+he$Pow5pqYUJx96nf3p-=K9#8%FR`HIW#K$cbE^0rv!vhj%(P>t7ks^U1F%+WCeGIu^jV(( zpP@m#-gw3g^cMPnSFG2f1ivgy%_mT*XLih+e?H>bK^&j78 zj#`2jWDJln@9*zFS1fHu zAP^QT${${z?E`+4$jS7^0Q+$xf2;i^^TPS1$@EwF_|j4bpgq4>{eu931_4O+TNE7C zF`OsqD@@`)bb0?J3Ydw6mmjq$nQ;y>d6}CUnwmbDG_>bYI>p z{7kESDqpp~3q=U=#zL&v-1LDi4!z}zqG4qq-$o;R)K?;bYd_tj zzR`kQ++M9EpWaM}VLwSsOgz02^uOr{4ddTrwHW#Q`7`&axJC5HaF!r#HKNJb>6<27 zDlvs4ZNj2#^58_sv$u_Nc88!h)Yc)|6{b=Qonx>-C*WSy>1vg!uh|Zr;qXj6KC53b zpKf$PKwyEY5xnTY;yPylAL1I1nqWGZJYblJ04k?1PUw-{&0Xd)y9-&k_!sj?>e~Q# z&B!VFrOg2X1@s=gTM-xgKr$Us8~8*qaEZS{@3fCQVB=v>WF53z;8@*`K06P)9a2+% z^w0MV=eN5A6mRg!H5R4^TQx%KyxX_b3Rv0Xz3pu{0u{oO^}6H3H`62Od%`2Yr+hhH zG6TLwgPGN)H+Z4@l)@KCS&Y`hs_C=Y-g?O>6iTbls%}?pF=TLFX?X$cX;s!V0jSnt zn4~yEa%hcqtsdW~w3y>mg-x|4lG+^g{{4F`suT7`lt@fJsLpV2z)#9-+~TPSAb^|_ zntQ>gdxPnmTDasxZ87~V+d&P2Zm(+Ps=)xvIaKEX#eC<3q0787Y+}z6I3;NcM zUu0)zH>#FTvRe=Td3C)9B;E?3CuSQ)!D;aJUs`?FSgC3%gOj7r{tg$a&57Xjv;PK;WK6I<#>JGLNFhfs=I)ek%NK4rG*Wjd)q>a<49-{bAuAz!rOg>G+eXG+a_Dib;4 z)5tw#^-tkvGfzs>is@V9rL`Uv=)D^ObP(q=Gv1FtGojA#E&htT^{_19tn~=Qxrg=e znp_|kzAbf0d64y$riI%vIll zUGDF8|6wZR*QJ!}rOtd{zHiBve`p-6hFM^bi8(%K*&H^2Bc&YtjQKXUpOnPt@9hJ+ ze#XXUDaD#MU6=Q!ilx;$=;D@7r~Lja6#MZA!aJiwpi?ub)}N95OeZKdh6|{HSS1de z&xRNw>%L_T8bJFiMrHTE#2VjJ6N=rAML6~38D+dFy!zgVUF|*M!mH2-qWLY%uN?>P zz|6Y(K8t_T95pX%?9|D_9C|6mYoD_8688?{g1?u}R88p~8E_Gb*pvMI$)lc`~$N0gLr%8niYza6ILiOqyZ}^SI=~*fJI)ySlD-E=K#iQKLYQ;(jL1P6(HOpzc6xH8 z&kGz`2J@=CyO#*t@&}?1f$T>{BbfjBON}iIfIWS`$oR|eB*(8EUu8fu)V=h8Rt+&p z%9i>f55X|fwGqfJjAbTY^ce5OO2Fet&;xtY*$hr!Nf8*2x3Cp_#DO_8&h&Pa>IDnS7T10$Eahu-0~jPXgki=jF)n=Sm4bCSa1BDJp+V)Br#7#UiDrr)NccQv|Ns zN6oK((fi6$kOycOUbjGiL(G}?7#9-gB`>I*_Fa$-eo`U+9n;oNZN4M;?b`S^DFAql zdBvu@52&oOI_tiUOwDrN{ey!gkrWk^oS_zm(zE6=VC|irStj^*AXW_{OaI@y00{5C zmp0!~y%Lg=kC=DD_4wKZz#3v2#P_ZFLYQ<}4T0RH(VMxcQ_u>KGt~;zdYBt}4UX^U z`7bG=EqDVEl=`jLk!?d4*x%B#9xeqOX|C*Q>ih|yqrs<7RxZofkU*A*Mymsc?Mp5o z?)VPLaiJ!|2{DhTo_M0TUd*=vcXC>oJ#CAIl(zo)NN2>kNx^)Qw&kA_&7QO6Zy(!V zzftvF(FWKX^52D_@R?7dfaD-gIvA~3YQats>CM%7Hu~_*$Hj$3%<+EW5O)^%UAm z!>EWWPjVd_Bc-!bERq?yr6&<>!p&F81wV?_e+%VEm-wD`(VOX|t2_A#5JI@tXai#Y z!2ez7e49XB08F|j3+W52)p=Ha=lO=wLX#;W>({^}SQ*;=L-MZd0bIE7!0a%R0=;s;t7BjF@K zYbgIYbKrZXoWt7!tPaO&>F-ls&@XyB>d4;!z{-II#2o3CFE2Sn^?**Xw=#);tD*ZJ z<{6E+el8Lt>I_B|rB$+&WXSpka5II_vGim7m%e2{&6xz6R};&-D4GqJk$h7&Wc`1l z=ZV{*Fw09bX>b*8(#{b7Y|1Z~*)WQ5{h)sQUWU6*Vu zN%Q($BA=f*CNYax{yg${!vj>ZD))i}iGQ^J?D zuk$VYA1&;Fs}QG~bQpZYKmJ>fDhAp6)+Yua#1KX-#}A(U_HAzhFA49S`+l>zz}_op z)2`u=9lEFxv90ryZP9M%?TO^OX9zj<^70D$p3R&=In>%HvWi%EvXBLqTbXE*t7w-j z(F_m~%Xp5Z=1-GAyCIjNo?3F0uFgIjf087=F{|(n9lRV|9V!6-d971?52qU|6xSTz zAYYRG&;-AUq$zSPe{0A9`!H-eOfT#)0be|73xPQ7pQWKrd4^oh)Sa%Yt#!GJ7>euZ z)>+H!MLPKBo`!sj{ark26+C~Ffx0YaI*<4mCzQ$%qxTBe4SSD!;ZA40Q<} zA^*=Jk|Age)a6q`EYbkm7Zh2hUqI1cgYa;?Khehl73xMj3mXX2(_u#Ajq&zvunK)> z?5`;LV*x{-C`~h+DU4#T;4U)suAc3Jg`%({AIXyOL&-_cvuCN@6O5Au^T;6|Cf*;@w;fbjle1%JWJ%eOulRLVo^pc`?N5QN`@&I=s?=y_wG57$j(nro<%G_oKAV1%wW~38*6I!R--Txf9^2{Wzvb> zjPtO_5}B7)JuWk&3qecbAoJ32g{8RHmJuzpre*uC3|eDqc75{DQE~YGrsZ>uP+Uz8 zO5bOW>2d14&dqOx;6qNsf% z1*?}lyC8-~{S{N-#I@2XqJAr?RJ1KtVlz$If21nrmfk_;n5;wNZ`Q`CMA22RP^Fx=JW^Pw)T+b{* zwEOhxmjg= zPD6)U{HIOeNx+HFe&^mpxWHJ{F^4rnx}{Yel8irzS5Gdk*NT-LRfg73MzRq0jip9` zOZ!g#Ay3sCrIt+o;L{4GxtmpWm9HIP+j$D(BU z5ZXW{bTQazU`lj&Q+)GIZ&^Ak-uokI-j~QHEB3sNHZ?CACP>2*A{>QO>gqH!Y?-${ z`Caku+(k&n^?0}MJ_Ag^qx?Ox$!-ZJBnMmOhlPD`z{gdW?m|MRrvI}K8OWNvW1^W} z30^8Fw-&q!=2o$`blXnSDRya)(MLu>{?^+}}zC95#Oc7j*#I3CLZmtFDgcJ(UF)o*SXM`4mmjH$Qy6Py6+7 z;m9V~3zmYvqrzGef$EvPs;CUeod7+M6Re^8u<11N>27kr+&UeTbvh1l?d0l_-6Ext znnn!v4BZ*QzmmCEf|XwfKCfpS?%>9L>V&4K1I zoPTuq^y(p6{BPga7#izmA==1BASd}kDh;SdoF;N|*X{}#z9dZjrBC$m&Z!E!rEGQi zD8u&xgLzfOb{>7$6k5Sxye--;5mf^2T)d0?A38$SwY4DzNQ2vTHshg4zo`yd3Yw-v zyn6$`_=|2lO84o#oC3MeE6^H!;fq+QSYT2!e?TMW*4=i;#_<7zETD0cRjIPP9 z<6oP6K7CnA=|fMhwsvyAC`+%@9n!dRPSZ&CTj*a^W)c=H_6@ErI^`b4iD4L_KV<^; zrI^F)_cyYEKlc5a?)df4>Agv_zP>zXMD>_;y`K(~0X=1V*Zcqah&0P-!aKKSK;MaQ zi>WzL2F3@#KKOYVY!bxBrP@to%5QZi*~bjS_8Fo1!PboV)Q-;z&m|c;4<#1_?=c9h zNg#oMgm-p|g0X8HjP+A<1DB5ir&YX*DO2mS5I?uT>|&AX@9F7@*iK7BY&GmPW5*jo zn^YVV=bEt=BH8ho8Ch=aw=dQWM!?9jLm+(KNbfO=vKZkdi64ZFujhUq7iG-nFhfu+ zKQyR*gb%U8AjaKr=J7p|GK)v|+%=Ii>Gk|Pe1c06($ZqPEiTV-ecSP@-)}U|v)09KY<$@P zg}iORx1n80g=)u~T*LC6@24}K??Yj6on<}NqU;5Y=MHQ+TCP2qP`qyxVtHf>;U35- z0t%KOM3`DuX6BsWw|+QeLb0psZnE#?455phgDj!{=|FgV+J+I?;NwVm;nzD*(HX^d zEQow0E@_J~G~H_dg~-F$Z0eoQ97dh(2((zt`qb*`Dvva*82j{kmjm4zkhkLH5d5Vk z*L>^RY)4%UwxiCD+_Aj>QO`4f0Z!Puf<92ZWDm>{)?F6Sa>+09Tza(RL($n?OVHT7Lt zWlW|x|IWei;GR72I~WVs^PpTEL*G8#$f!*zdXsl@4P{Z=WN6(qZaTe>IaeZ8t zAPL*u)kgX{mTK=!ukt}mHIeiR+7^cIthMOMj{ai9%#@Kg#|4=BH`*jI>%Wzo>=2D% zsB80uqfLk(xpMJe$(Jd_!eo?2^5i3%q_6UM>uTWbQxk9oc+$H-U-2+WIK_xGB=^-L z!sh%`>7qccI`F z*FZ}ahuph&57RIC*0(B3&CA!M7dPpmt&5i3x0X8-^6Ihs*(Jpo&a~6!(P+1+28fnR(0n<}Z})n-$}1cm=2+{`Q3O zTqf&vUB#oS4ZFrn@-(jBFp(uMT;U*}SLowRlD>JFgO?Z0BiYZ;m0b_eB^f@1vauR- z4ZanI3wcx~V*43>5vV-d&={v$$?m8Wk>v}`F9?7?N|g2v}tbA0GrMAu9IGpNu${9!oZh?o7M3HkSn^YR2>o6?hVIc-BH{K2YlL zaw%8-^krvQG`*zi82VDXKK=LSYI4}2{IU&f4KQ9Qn>PMfFmK9{sug@d{IbWaA76^v zhqets+ZSe>bZjZxCn-5;rxM*LxS^9IBV4@Yx(+xVHKPT~*BZKdaK?56B}H znAtNI-_IheXF;$&>Ayv)_N9zwiGST!nATvk50M$*o=O7#-yvs`0IKs>H{r`K`UUT? zKVZ)$x3;fThD^@QU27r&BwKZ1xJWO2n4BR=kX?HEB{3|m^Wjk8E9i*3$!1b;#2_8V z@D&tiSoaH|PuThf-pP^Rs~>EFJM+ayNO6>_VkYw-<|-NG1Bi=~nw_u6x>(e8VNDBX zEMqC|PTz9)x7KKW&Pohj73T9jUx6?e(nngByj!F?6uX`0ZUZsBIoAC(Y{pUF5P9CC z71B!gE&E?AJf)5oq5r#C;=%Ar*6iKD45XbU*>_E+fz?(tV_+tK*^Eywq9gLzO8s>4 z(Ghp`_|2%v&G(3UY+K@pDBMN5PXHKaAe3O zaHF(_Zmp^wL2_+ZCj|QcH5C^-ezK60BV=G?Oft}LRzgOr`VHBD^-U4R>f`Nwet;bT z2i*z&f~IYVKCid7mhZ?owBXQFMEHJ8p)kP#b_1j>(SP+Pg;-T5X7`oc`K>IC{Y@sY0H8Bk*51Lnw*&-Ajs%lO^-VhPvLbzYCuN7 zPrp{81yWS=1c}b9C)uJrBpp)Epve2UP2L`NPvRBhlmGDZVGD!6v!KYz%925?vzHzZ zg}q_s3||;uou~M&`!&kZu~g_lRL*2`mudtpIp?=@#isIFAj0e8C|dN|1Tu&Z9p(E7 zr4GBWs+dFrG;M_!k-3?1O|$q2_kmHt23UX+9hs&V*>-n68#>CRpjKPX{>0rHl)h{$ z118A{^$>*g^w!~j#EX-!l}Ov(M&grcwC;B?dYWZ@qK0QYw8BKCfwCkgchn+it*9)g zf8&@VZ*?$7hMy*`nIh$`@mpB5+u}Sj<2uJq$zlD<<)oQGm=E}zO#qRvTk&yi%oNyl zEZx@7Dyx}`Mv1b}dwAy;6jW{VieSdK*6%Qdj%!xVKU??p}yjJAonrPTN z%WOF06YH*>cKQAp=)By?J546-gs2mS?+WS4jT-#Hf|7}ymi?c!a{einy{3mHjG2UH zYklUg#P(opG~l0~a!y2RRQb9zoJP)+of%izc76s1xg_{|2=iB?{Mp-Vq4K}KC{cW4 z^PYeE)k~!Y70GtIV2)kF=@p1wpvz)}+Encb73;c;`jR|bSerEi>8 zc}oFJJPcX1^kEaa!>sylOl~L*>E(Ji?#cN6m||k@@tw5XxJz%&y4SLrtzVBWpSF!8 z`?Hh4`|yr;;hw8!GZZoF#8#L*emVa_*>4_D&YJ}AYHpR5n>>SnYphrQ!T;=p(_Hwf z=(=3Zu~jmBfJ^f?*F$a2BJQY8YV`@DzDAC2+8%F|_(Rbudi}9y+VJ0yfe75tDh~zH z{2*q+GQt%xmd|>A zOlc1FfsGtCP;|7sVxJPFn#$nR^}hH^Ie!OrkFWlwS(XlgeOPiPy}uw4Fq{y5m5H3_ zP9Ot(*AFO@{!55cdVGRBG8Zgl6Hq^D9>iSuF)Y)6Rs9UX6Op$gtWaW7DxQ|oE<$KFT%YR+(fV}^KV{bh!qm3%oi11_SzAzJtP_PLT7M| z6jJ^{!cejpCkY{>DGGqQ56Eh%9If=(mnjs6`L6DT{9Fq4O%ST6xNuhApPjQ1^9S2W zNe;O7VfO}PvoDnNjSV1hv{{V3PFb`avv&#kvH<*Ik~A$2F_qHIHe`F(lK*IL=-V+onhwvKPj~E$XFYRNkyUmBnKhq$wI_7gpRdxv(_QwW8nKzS8CL>`G}DBFJLtkJg`|{%E|98F|*Bw!s}v* zzsYcT`rPN>k!C^rHiV{?GvX?Q|9sxk{;vW0J;7v^rj)#iIDg{DpYbzVfT1{(Fj3LD z2zztGu4;aH`osTMYJu~#P`t8cpQ|*8a8sWddCDY z>q%+3#fj1ppe#fplg=~p(@6|-e+bEVko}}+FmRW&9`tL7o%jZSr`mkU1uAO-P@pv3 zyl5A3950cA4RE2w_~goi8Q0}yt8x2>O!JO zz=2{VEV2#r;rp5S$q;{ZpmL6?6Avw5IH98^0$#_C4h*hjkA#m0q*y=y3L)_b&24%h z?#>oO$+m;I_{1X!wP`SM42Hlq9&u5{EaSoC!1+XXmuesaQPx#oqnNRpqGc2P$$dQQ zPU{L+@)q9`?rQ`-GgyqUVuNj-LzT_1k;2pmh-Qdo+{irFw*ofj1%l;}^-unf^*LM& zQ+I<)z3p)VN{_n)KW}HjX8$|Go`IEDOFqF27t`k_M5Xcu?CUoPIC7q7H=xjqKj^9+ zQlILoZKEl7bjd4%G{Ap2WDaoDH$JtlohVP@%UxDAe+>eudfq+*SHnw{1_$yzc(Cm! zFMmY$a|T&BG84VbL-}x^&*hJD^A!8y&$HKfDkOeJey6Jq`^kP4^Due8f_=CEyOlPq z4t~I(!$$saMIJXM*&_T*yGEr^9706MO7KnoW3x|nAge;+o_t?tAhgo5Tx-1dEw?Xk zq2MDr3=lx0e%&XB=c=Ziv=%Nlm!=($yQ;+5#`VKbcaK^&q{YxModA&9G(gHYR1F_w5b^m6A z;Dj>FMUW7r3M&XY;(fVCkL*L!%|)QPPPkXG>?395n;_1H$Z4|z$PD3#?gc|wD|S7J z^3P}Z3*;#4AKTFO4h3bFcpSDAm(cWIxyU(U0tona5+D+AwET2OTkUoze_<}bYi#0j z5}EEF3~7APcA73TT8Bb+2*q}Vt$UJ-gC*4DHZ_2JbH32+cAtCc#wYTy;i=4 zOR5-9hw%^ua`dji<}~SKi7G_?KK-SqVUwo2hSjs{MoQLSnTVJC%Y>Echz=nQMK2wA z7E-Hu@zX-imfnk49kgDM!G8DZTRkF`{>J|$v(7pPDgPXO~1v%q?|bp;*%gz*%%^N7Ly+S4?IFX2^u@uM zvhVMIstJ+#D!>n+v(lv=K9lEryD5!N-j_io`*xGZ>q3^!%7-0&g8fUlH4Jl{(o3uw zUtSO$O7k&p*AkM}N1qTzk#q;@pqy%Ikl%FwR;1BH?$*@RaU+_pePh-Kuz#^eEA=If zCEwA2)RX_g7Q>;WPVX@9p`zL$C?TLv3N%V4JTZghA-xXraOzoq-B}`VO_t5^9hlll za1QIR^rl&6Yy92ZwVmu_Fq%UgCIN{U!C&wz9Chw`AF(`=ND zslgmf&?7xPJzjaO=k-%5RbZnJlU*h+Q~y5}fPEWE>36q-U))EKTe_HH0c+THpOQcDb|g-zi@QivLWqg9oYp4)=cOo8nwO*e5! zkP80vUTcLOQElW;LNT_-E(L5eiwCaN%-_?SoTziYP% zQgv?(G-gfIlZwFbF9T!<*+~L1ICK$M_f7kFkb`U7Osh%{SFGVLiolir`Ai|S&dAry> zlFNht+g#uo*~)YT3+C$VX4G+c5r^GxI5o?nhs_Zig9NX3Nf&qvQ-$)D>Rg zwm9;ZqLqG4JtUNL7xYU|PRt4RNFz;XMbXZL24f z9hGevV_B$bg>~;6DoBZA{C$zoERofF5Jn##pIn-;V~@{(+XH*Bm6o4M9$?av`NpP6V(P?x{o4l~<<3DDc*jZlAOu$UMc}7%MCz6bJqg z+1fk*0BMI~U+w(eNch>#v4G_FfZq-*tgTpku19;*;kmxbY`jTI>+T{bk z5JbfpVOT#~&DandbFR`pBZ^dwO=Nross+W7YJS6rY5$O{@1+DCF-`x*N{_O~L72SH z3ch6t1s%*GU^8lY`by>xTjXEjM=p50p-qql>Yn+343I-MGOjh1O2dov;3^ev@t=0o$fs6qnO(A}>AW`|*$GaoHx_Bph8)R@^Cg zKTp-xBZtHa1EkUw%DCht_`i2Le+ZHrLSSKm@Sz6odlwe!^p3>VgM0n&XGj(w`l6#l zPrh6j?+ksuKizz4pcJc=A=TOZT1Zx~c>0Qt{_<>aCvYL>{BTXpI2;>x^&Vv|AR;fX zR&Bb7PsquA>N`clx@eCS(-C-b9mD4%N}?t(OI4FDJPPVl^*LOOg`4kO+F>FP)7!) z%2y_qW1IfEMg#j&B)8WaoX972;<>G+<5i}}C+dn8D*1tXciO5n|KS7(9atqhOgP7dC=jdi<~J{bSVnEFkneL2}uZ z**xEiG8d(}3Y~w?weZ|5=Ilble&X}9Wfq4>QOTK?oXf5~KpAe5K;*rKsvx=}yi=33 zl-$_sD3!)T0|Kldvm>cHsaeXGlUH>{@FVh{$&!5He}=G({`;&wzrGMjA;@vD|3>$3 z$q7k?4(ACt*fSN-r1PhO`rVG$!SdtC!4O#10xU3V6VG9^?Ms*LtvwHqnjQ`=g2?dtr1O8 zr><&cT)sx$4|O73cyYA-5ug+Bv0L!Xdsr9uKbc}Kd9|3qGk|R~!g7FC(Mr-w z`T!dgVwy-v{z4b4)tXe@&_^7Ug$x{h5hAN=5P49S?jydFMd7{#HB246{js3zR}57$ zq@TjPs%LY3wy9o_XE6Kl_^PW1iVo--t=fGmu&s_0-j*7(wnk!6a)iPLyey8i*bWJMI~wHvb6PB2a`vn9;XCdu}de5?)k zNQ#gNdr^9=07m(Z0^f~3fWVJJ(UMp}>1-cj{u(#)Gu%YzFb`6HpWJ`{+`ILaDh`M0 zS;*X)-fXCh&ScFs^h{GAD7GW#Ll~4y?qU2U&aBPL5*J0Bm19w_H;pGz3$&>jV~mry2jn^g1xp?S1`qik)a;j zs_B2hWvRcXD@(*6-#{@5MkdlU`FRQZE6E{yR^9F)Pn=pM=Yu?jk%_k67*^t6Qwj`k zHi}jjNR!VCNZM+roN5=e-r4Y~C%dIBN$FgZoj%pb!F>)D%iE9|WY@_7bnQ)i#wo_O zir~()E8w&oRYc9lm{{lJyb4@5{`+xq|1|m8>@>D+yFWc;`f_XZZ@A4MJ61O7Ljl#} zG7}QPs8u;a5gFn~uaQgS0dK{uk|{X~+%&hL>XK@7k!v^gYazrV-hZzM4+5#p=jwUb zj!YzE&BH`&xwT)3@?B79+fO(rQABV;$dms3#)9Fv-TnCNE5n~CokckAsE=sZ7802u zoa-0j_P9&|j%Ef`?syrY4F^}hH)+svrYWPpt8b=Y7*P_M!mEOfCb@od zU&j)ob}*W{UXG>BvizDcP2-XtN|_sh%wq|Gq-a5isfvdp zn^I}t`48kNa{+Ve>#5FIc(h883W<=b>16FJ22&26%=ooSyswzvZpJxHhjdYWs1|Uv z6@m5TtAGAa0e?6{og<>G3p5mk!nlZx1EQZ=oFKiG9o4gt2D!%HSe_*s8&hIT{L}cb z`1$mSn8ueAz`k1;1n3McoA7!Y0N7p-OmGfjTJ=G<1tV07wG<4fG_E*pzz2PT{{@39GA8u71^^r)XW zV+^bfJ_!`mQ?6exT;*@Jp7xwwQKCcFqG$ZL`W`K~LGp}~1)Gad$L5BQ39Q9&)Z;dz zppa!B30W=(PQdld`?6*iqlmjVsyA%7!u#y8wt2?kD{gqlcZW@Je^L76cLpA`_ z8+Dq&uY67?HgEJ$jY6!Osrx>ttsUrDa>1TxNk_zXBagkHk4sX0=hbHkMvtT+CG2BY zC9eeY@H317j!goE`mmn?IGmlwHp3Sj2>7G!YJ35}vES6EzuqOrJt7U}oi!EC7dMcm zMyILD*skdDMp)ZM;!?I**Nd@ih4E;&>76jT_XOb)MKtz=uIn<85Q0UCI=0`5|5@@l zUO&;u>BOw*qv_f(4topjH@X^f8T_`u6kLJaHq%W4TMbxkIn!ESKkTPCCY&mV2ahW! zQH^XhIpcm=caH6t@zu6)wn+)@c#Z>MeU@BlsTY$I`RAT`?@@?)!{Hpj+2c)Q7| z`RjqrC56;utsqRLpO9jveD%e&0a1;NKC+qxMe)wU`aEd=(>IMB{ z$(?P+$qQS&xQEaAX;%`HYzs&E)2xka#-nLO8)^kMjZpvm*7ps&+g^|Vqk~MJWtf5q zVHO+Xab)Q++l!&{RB7zCV!+2nnfrtkjbB#4tgl;f--uJhI>V|%$qZBo{mCWhPP+LDSKqx< zF&a44bXhP|_jCz7+MBbxaI%Q#AY&j~Ct5F8Zi}~91oi3X52)wngrrfaOhd0ouY(?^ zC?%&Je1W>?(d~~hvsujaF1l863lif@h$JLXqA4}jXz0AD)ffjYx?O4l1ygu}Y|k z--O85&fwIH*FklRUGRb7`lDvt)v+|x+`+=|Ri)FIofRepN&Sgr{d4kQsygA0|0G2B zjljxFY05u-?rw9z6@ji%qSkf|7g8vR1Q(@qd@<26>nPEr{h5~(z@ttJMA2)965H#W6|#_maMP$2Kw3ulUqN$K6|AW^C3CRU4g{% z9!q3utOK(Q&Y6rgr&syV3G?uDD8B%gGJ2!Zj3`z2Kdeii_6WR!h^;(Fj~tCc4RSY7 z`#oRXYy?yERH;eHz1H|*hc6SC9#K?=vY3bA+D%7E)zZCI#DzpFaqT<8ySmQ97s#RM zs3G_Y0Y$FhZ#X0>BISIimAJW7fvJ|hUpWm&yizmliJ%Mc2A-GLQuc(=&*zqXTP0du;EtpBjG=~MgrFe0`s zs=VI8q8S3bwmFZuT=g{M`Y@7sD>C2PXWZGERV>0V^J8Z%YXxejTd$g&qm}#K%rV^u z!No2(L<@P5sa^9#!O_=yUB_`W9E@X$ES$3@orB2)J21lZ?d9w|Pk2+f<}ka4A(2lM z;Y1qaQy`U6ip@`qjio8TgJ50D(qN{ZWT8Jna!zXa6(~+pAm+SpWh(#Du;IuN@6Vf> zQW!TdkypGi!P4o!rV}Na$L9@}Cm|{$SKB2o4TKk$v%?=FLMAa+>1<1~d|SjRZ+SHdxN93Cz>-9(dw&;Lig()9!R%r-!x00RHTJUgLD_rzZ;g*=?>#D`6*Q%*H?4JB z>`>q-d;ryf$EpJblL)97;K9xQ^X_0Iw4FakfdMi@zR0VvbbTWSE$-N48-Li1RnMra zN*~oi{a%xeD8}-!x_QQC9Pw@)oey5}gCT5DVL-I^L(I(^JZ#__?1)B1UK>uzn6~!+ zfPkR|`Bnzz_EP3)F_(dbqrqi=6dUYcH#ruBwE9UAw{<4R_A{Xz2l^qkQTK!b-ZSq7 zO+);c-oXEn_c>CN#G0|WWB6cyZz4%&Oga;x4Fu$SdX=QqkkkAKYF{w3sPA|s0=sj! z7|5B-)g>agrdOH22j^`34YFFcOoi?3)^O9%=Yrk&L#tO7vKa3nSPj{)$Yh8pwJCEI zg{F-$KiJ6n7+&mFB!-87j(6SUo;LrK9vRjOS4qPOR#y3DrgVfoA~&jO{l~prthR^i74or}N82yh33tDs zkr5)cT(EOIVn~2EY1Ah{|Dao7SS0F%4Xr?L729x$+dh)qCZ(QCP3?|CKKLN&jKiVQ z$+Lx*6Xy)a-?*Tel&6)ulcn*E*P3ZJ=VB2~s8gP4mV$%V`U?86A2G6Ob&DCkELAvR zoWU@$qr1^ON-s;3xX}I~IpOAZk1GaS!{|siGRf}lS7cTQBIC+QqF9T*NnxioJ3*+2 zDhB~}riIjIwE8<_#r!l%WR)6V1rzj>VCdf*cM7JMo3@gyc$bvnNszotn3sXWuqf9mBC5!H_Vs(lU?>Ikg$7Y}F+DBYsK>U#CVtU#uKC_t(49Kgq!Hh??)wj%y zcVSG9H{0$^(Y^KYNH-$-H_gY96V|nqg>R*&6*Z2d8lO!>HnA4izq-$?OtgL|Yo1^q zHIC1htBBk%C_T18j;iTJqpsDFNvf)UxX_GO()Kgan_R6{$_vu!+b!cFxW8(zct38y z9b|DTXo~58)f`ilBvm#3Pqg&-Dh)*mTsf$46$1S|UL^EZ(@LkaDm#!7#C1{eM{bS9 zz@YLQx{wcI<5(gTx+b^p&$^;Hck#I%=P6q>!I3YjX^{$LgqiI~x+wo)A#9Z;zQmGS zaZ#8%h7uAaFItpgIwg|BcySi{S8gyW>Ma3SO$pcs=F`e9^_j)dCf&H|0q%J`TAymn z^HA~}=!*M!+syB1pAZTTZ9TTals6BVzI@#ydj@~Z5YIBUV)7HJ9+sn2ZS$DonF_-l zpQJz5=I_Moeo`CYe@K+>K&W-u+++`-`x5=m1ed42ba@&RJy2}qB@39QGQg2hTpd>T zB9;?$1al62SAwAMLXI*NZcj(xkvxEEI7Ru?a;C$R}Tx`p8SDM>Kyy$vAi3I%h4l^_owun5Aih&8NUs$ zFnrh(NOIIU8DtG>9Hzult!t}l-XM!H@y|v^`1m?*G$3^Ixz5c zD3Vhc`)v11R}vRp9NGv>!iaSPU|gE@GOa|9sifJ(Ku&_c{<& zbd*C$;0U{{^rhMOt3`+F{vE;uC0}U8F_mip<`aIuJJq!J~uqPH{Q5e##r+~D2(N;V_iYjUw$=2;~Ug+GD)%Zw~>;h9#U;^*Pv&c2_JtffYPehe* zqQ4T@Bfa=xW*SI9oJ|zFCmc^$gu1AxsPGeK{VsoK#E9-bk!($f>J#Ty(NIZ$y79?@ zXw9Z-${mE$&}kw=`6oDP=|lm{+U>q$y=wJN$RY$k>?J@jboJMUqD5?}Y_EAgUSE55 zySoq;C|eN5(F(CBBhzQRC(JAiHOSPCWD`8TNxm|CN4`hZbvXE=bIQpF17_d5sYcRt zeAK+Xyh?sx#2`$CsnR5Er-^U&T@hJ#ZyR*vzN#7eE_77MT2+L&H6kO z%^FBll5!o=*b61x5r#g+-}Znw&Ig~)PE(F>h6JR&xa~rKk_x)i^eLwkFI#z@c>c#v z>`S?^Cu=~K1(U5jQya6+!LaE>Z0oE?|J^Lxs;L3^ZiP6rp%S*vyhq>%$TmBqH4IgI zjpnO+$~cMB#tQTX+o}PHEJ*&_jAtIxF+V!Q8t~n;`O+G1Ga~vSrhlcho)0xaCuCK4 zkLy63g_3ASb|=OVka^QrNS^*E0K|g6ycM1L+w1ChV!jf55I9ONdG@nxm@4gY zID~uyOsDx_-irWJsX6HUKr&tg@B2}N<#>tlzLSK^+l+EBrIwTcl9t%}E6WCKt8~1d zo`73G*7kY{B$LU5h{jv*fAEpyv7qpkh^pH6f#PRMKx39ZJj`;~!o;5p)UT zz(KG7E*5tpcjsEKk>Wth#n{WkW>aMh(}8njanM;ZyZi6Tay;|}*t#tcCaGCiM@-t) z;~yfG4ObRS4DFojng(vd-Qw7bvB#jb*xyJkm(8Jth8Pp~xj89+jvqhR&) zZayH*VHVmWZzkNc>r!=HDMuDpae&cvml zJGvL%qcjFKYE2ZWx{11f?Jz-vux;ZVqMGqQyQ(Qk$aA2?JmUs#$AQm%{Ek+C!aF9L zV{~1La~b9AL-HbkgQ*6>Sq!GBzN+f*Sa^(~^!PNxzZKXd7JP*U8zPB}k}h9#|1uQn zchW?hyH(lI#)GV*hSRBAJRi`mSQ=jnfXZRlSUrB>>+M|wF{%K^D+TQ^!Zn>8j&=D1 zdQ~rum!;U%9S0fYskMUgXTXxnn(@`^ti&574YAOk3Suk`^%yvSqB zkqS;-v;diRd9uQ)6*CXeT612&-uRxLP;ou#NtL$XdP_e$I~$y)>wqoBnyFD3>#o?X zxQr){{f$3>B#s;zALhkM9))3LrTLwL^08js6zC6#%t(67^_DZrLr%mIG7tV4TylD= zAOR4?q1)XjL8D~cnWkYZxYcXbb;3mZsm7H|)_{vl$mN-8;h<#b^<@IkI?C~rs|7ZP zDGz6E6n{F-M>ZwBi@Fqdz_V!#^IZSk|K*-dQh<2_7!Qpv{g)A9yQj}r|GV3t zc^~wv=$!POud8=^zS`=>09!+JAvgKXprQp);N3F`B<9<5H+%>vIL8V#(_KENW&$$1 zVf4)2#l;089GM}+OPw8PJRU$(nmC3!e1DF8dqSZ0n%Z9L0dw#}jo_sJhA!W25nzF% zh`BtbY^R4iUerfL>9Hvk`6-PU{NzBb+@ic4L*a+@jNZvSYJ0H zX?tWVKOiK969%}hkPoSiGnFMgYf0c-rvLxBoC-(Nz|=^hWssd~5SY$r?ZRvW>GCa( zfHJL;Q^y-<>)?wIp>fyUuK_AYvwTJQQuN1bp8eVn9#2oM+wh^D|M|j{iRO>xvQ&{E z%P4~=AyS#$Z8zeT^<@gtOF16 zE{uS6?i}zG5>YH-T}@wWuIHh3mEs&mfp~4}(C$-7i72m-Pp}T{UujkWF({UMU1x)N zWw?C3@DdzVRA~yp_ipU{BlFcDxZfCDJ#+2t1G_R z6aVM1tNMB*nZOt`q47e8Tq`JR^%Et&{s6Fwe4qQ1sU(2-{(>5 z-DSUh;qf4qD5?gjG5KybHm@;ybkdonIW3A;8tVjE8K zec}lgtuJ^Mg7}u+GVlL-8>ak*Y!8gsMJC6y|;qkyn7ZXUR82pdTUjEsa>mn^y_ zBmvdp693ZUyB&5p)2zVnXC|A&EzfTWR+HsVEs-p5InWd+dxFv$v&FAn6ztIytn%HJ zC4!JvRe&k8-|hcv3NmpmB2u~d9+*okI)bW97S=pJH^(yRz!59pSo7+^W)ZXg8=K=w z_J*9Fg$uS@^WaV&HPIGG6DJQpzD*yk1fL5oDqlm2w}33vKiymWk@o=Ev2#AgV1=Qk zbRBJ*J;=3G1iDgf#KmMJ>UST}*IQ3+=WMcruIdh~SshuGAfj_#p2#8}bmq8ocJ>sx zRzgGl@!?03KaErwuWge_ZdfQks=h#)kpWBG05bLPE!P0(OU<6dIK5^Co4sS{JUbeo z@F-|Et0Hos5Ck|XA}Q#t68W0#TtVCxNKHF8a=g;PEzUFtQNr|}YFskFPbM)Zhx;oa ze|;Pm#&2VWRNe+?fo$}5F}hwOOV8?Kr&&Qhl%jF0Y5nThhpG}}>inIj-3+faDoeWF ze~uTi*`x#9avo4rys8kY{^={0w`wn#r+j&^-0Q1|8Wq#Y;v9MmkBKb{F5dy@*h8mN4;?p~96ap_pv`T$uD9SO3+uot zI-{81{juLpk#+6jk*|K)ZHNO1Rz4uEcfpiyXx4x-`+uoi%i4ZD$6iYbFcJ5RpC;%U zMKOE%5xi(rMCIoBGPB$G!FoW9?=cjg#ui`$4M;1y$a)8w`N`o$ZgxRDy;UIgf8%gk zy2Tz$F5CD&f9Koe@J0)-P{MEt4!M!l!{!@Fn0St%0q5^Z>J|v3Q9j!{7pHHiMmMQA z)*WmizlZ&`H0AV$2L$IB2FofczEm`X2>(fRVi`Y+OdBNTZ%!DEAm))iH#}_}*|Uao zSP=JZu4pOLmP`=`ENTqHd02QC)M)GE@FqAU0j|?-UxNLy;4eJG*j3(nK#W|h)b(KD z@E2`^WIIqR;t2{Lc}w>~Po96o88YV0^jJZ3>|)uj=qP0NHnyq_w_>7F%(d=1gj_SU$Z4z!MGr+2_$}Cb zTf&23o~P(zN#PY7E?*WRK)VG`af`oW`52Ed@KIp#Q z)vTPV(N~iwStWBO*@@1{!KSjCEZx zsV-*wZm-H*#<*9Bg3kBY*LqwjT0fR^HVpa4(;I1LC^2-U=D0oSP08 z+hiS#utaPPJAWu5BfDIVVoyC5W_1xmY7AOi{gEi+=sixPmA<{I@pt8_czio*B4jGj zYn9xn2Gnv)QB~xVPh)SAHM#AnTetk}vg|ICoWIyFI(m)OvlJ{cN;D@_NR7nw>M{r@ zpEOBb&fml^HIJqS9k&wP6XNlWD?*mAEl5np*k&OKNQE4Y@y}tQ~im z?+oAk=PExhb~HH97lz+jR%xb;*h&K9AOos*PJEtx_4?#T3svgzmk$_c{xxh`aBoKF znPI2y+udQ(7HsC7-7nT-(5bn3+y-mVYRokGWQ46*CsqB`UJZSB!_K{Kq{<}q0-oLx zwY!ozQBNqLp|7_5Z$T54n`z_=&4BQ3ioW>%%E`j?FdJ(}a`254sp7}*5qnTa_RHXg zM@B}DeP#*u|Je|Bky+~0pM#Tji?{VQzep^`%P>E+;JiuJ++=)>{7L?GlYBCT;V(-k zpRsfv2VpjDcVmhNVK%5`m1D17nD5MNh(UR=!BlP~B`nOqRlXCt^61?iJJ*|1JZ_nu zoO0ol>e}E8ZEdqey6;E@S)>l;@v$0ozSuNy&76_2tGj9vx8gy#?)Qn~<;x0}2*gt2 z_T1&vSYw08juxJlDb*1%wm-*;xlXG3B<8RwbA|lW|G8Z+w0b8|9V61VE|+1NVJ!)X zg{^5s#nYz~!?pVHMyoq_EPfq!??$wXQbkgM{^Pzq@Hzif4N)(|-!_fb;9^I~y)xYR z#U~7Eo2-xSJ1^2Vsx5vIjwiF$y0~YYDe%ysg#nzbw1)-th|R@#xVlLsz30a*F7XLh z+ji}fPdvhvZVonwGuRa(8v1t(O=ov*hopH27R=HT~9Ot+8v7OGO@zJRiHIK|LXnn?vK!9fKeD0of4UfBfm6QRNDW#r&eY zrh`S4o4-Fi5SK0u?4^A=pBnoUV^Z)fy#3ZhAkzuPfr>dJBnr9&TC6&)a8_c*1XJ1y z=sLkSZoP-;ojaiL`ly{O%c%VYMppiE9SX`&m&`DAoCy-Gc@~=NJv{vP^gDZMLx#lw z5)5&V76h8`oc|)|^FJvdaf-aw2DF23l6dX#mUo-qHVnzjqzGxLSgBbdhT=G!-lgzj zSOY*79LJah4TQJ<^)$Xc@t*72!>sj6D+NtVyLv0Jfw=DFo|A2w&v^@DkEKDzbpp6n?A788X_POx(l1&=|-Leq)HVe=Cx~zZi!KUR*H&~&6igq$WgdqWbXd{0HyF&3&w|0su&zBENd_mFAI%*ELwQHi&e{>B&ACu&}%`ynS%C>M-O5_OD!9AWLvYv|D_( zuM0d5bZaqE0#9`!p`U$1MfcerZ6sXJ#Lt+ zk)N7$(S$QLF-3-5EA;jzhV!oYpVqcE9{mWROW}X*pY)qcr%Q_eT4Gd?_^`M4zCBgI zVUL~IMPu94PXpImCX#_y+&M-cwRBhe zK1a?npBCneXBlG%qM%d~L1=F!J|f6X*oG>{>RPlth;wzYsj`Al&OzSwCoV9}qG@6Rxi+J3lMO9|c4_!yTknyV<_Z>!#GL<*F|jO`jn4_7Nz zzbvwbafgAd^&r#2H}OlpkY~932L-1=0Vl3;^$8Ww>p7r(w)0;oqwexDIY`WrQ+(CgRai4m6ZNr0HV zSO&rZfmu5&pFb^^@AO|LP6|Dz1I%uT-jza5~Ti+X%B6Bi8pQ z8&ULm*qH~JA|3Po^lV6=1&UMbvKG2dd2Uq`)Lf~ASxVXP`{t_ep?3*6-oW_Ro3W25 zFOK|*3JULU&>Qeg#j{OrH96ncl-2!{ls$2G^l#@8{k*WcG{QFO`sM>@bbM6FROnk( z6VFrLy7C%`LmfTkq1ESFiSjy|r={|0u?Gu;6ACSSPmwT5>b9|8x2@mX9=- zz}?)y_7zy(CPYG7ec$5OxLn8)prT!d5W+NGYFZduNfE%4gudFU%k*(Z)vlqz!QiqQLwH80j$D_>bDs9Wv(biskc`?MaQ^KeX>$s96%f5a>4 z;l-#RI);4dLHXA?b;POCdsF)CCkOeJXBJ*}yoblGDV|O{Cf}lHQ~XyGbCzbQC7#DC z-1fE4FWaOkjb*uOkrvOX;5`EGWxM(g8~FoYVX8@ITap`N6E zL7g^1uBX~6uK*cdD^PeQ1UdSrlL%M_zRgb*Hw}Uszerw=Z(wB03`B$v9D~8m7Zcic z`F&f4cQ{Wf+m%0i^Xk>Bq^u}cJBo&TqiC~#e&5Av?*URyBO{=`4|S3QNE6yZ3i*4? zP3_3xY#6PU3BQbw?kDA|axIF0jSVpIv-HF7s^??+>tu0eyqKPg(?=RuE&=F4@yp&M zhO}8JxD^G?+uvSOfxpyiJRklZb@spLtQB>L-r@yyf-#2TE^_So`!G*afE8--`IsWSbRf+#jgdgC^4p8D~K{%t>Xh& zl$-Sf7)QsoLjnsh@%2?d&n#4NauIm|XB&yJ{yg&@*3-oBjx>tl`12>$avJ<@==_ z+8%KEgsah#Obba2dh9N0m_>G%oB`=C!zZs_R>)_?X-0}MvA|E` z&gssw_$Arqr>|ryOR}J=4O@Y5eG9T`WlG|5(ZNKq_WnIw^x)-zqrrzgV1|{r{>?w| zGvy$#4=hYi-)Ik4g4|+ix10@~O&JJhrYd8ZEcWo@Q}Qdd6WYox*ZNqIR>S4R9Bgpy ztw-VyY|a`?@^!ZG5boc+%Ehu#M=8N_nq&dPZ=c)Vn?t7!G2>D+E#Dc}`*D3UD3-;dL=n}Y9 zy9_=w3|x9WW_+M|_i(kWmH2`qiZFZ3y|MperC7R|jK^|Y1RWJl6jVn)O}BR+Ugh6f ziRe@Kb(#@I`Mo9H)qXQATwHBC0yppxOi@cCb>&~LXhmag4Ki3WVJja^Xj9MM&#*SN zzM^sTlL1!OP{Y;+PPD6Aeo#Du4wE!F+uAe`M!TQ80Kq0+%lR%szqnlLGRM^(1xR(e5mmn zL?vKnQ<7kPwzl*8+VOwq>1dU71RG6_(PC&h93RCo{zs#yx3v9det=BYg1*pMJ4aebPSU_B6BCAn=;D2TY#9;eX(9^ca-Dl(%1g`H8CmQu2{7AYAH#nyo|Hb=5s;%R zUuZodXDa@M9MQdLypS#-l{N3)G}g3{wW(Ba&toBCA*5TUY48?58l6n|XNJ9MWxAzu zuLC?o69}`7KOS6?ggm+nUCwvRQ6=+X>~`~_(R%o{5_;q2gmy05?{ElNX<3rWKh*@$ zcgrZ-i3EPde9*U9uJ7aTS$?h&++8d+E13ffR1wKbO3xcdAHJo9f-|?pS z`Q1oFM8q)TUw@4(he_@3g7Y6MOTX7vTY{$QZNl?v0l6 zG0PWKoTEsYNHj;Uv)o3fLaQJK3h3396=kMq2Nf-6Y@;07gSU4-(wJi=6I4doh{NUu zmi=Kh$CSMedhs(P5GGh<>ks9BvM(T2tubmF%df}|1d2)IQGxzBe@p&oV)MVSyrbdc zEB!BR7jJKKeXX4p;2sMkRZ)g4ur^Sd=0Ekb=jC5(Jd5R*&(HC9T)T*>mB{TGld&cg zowaf0wr@X&9w3KfP%E`1n>oL>&DDt7>X=*wbj6XL%@c#u{^;Ef16xk#hc14woXT2f zL>|5=*^FTrYg0x>ZA?Y<_tKJWiR-RNh2y}tD(zcP25Ut})s%#l{IM_hXw2sCO}&L2R9fF~xeuCA_@2+3-l_^JS;m7l`%Qb6dZUUtW&9Jgr> z?BrGqvn**P38@3qR!Jqdvh&z*#5gpk##L$#K+V%w>b?Yr*sR-Y3@-V7+$UT|mv&>E z*~WL$E%3?R?}I7M3GEi3X(}8S!ghbw&v8wMP5{LI=QF*r%kdV!gx$9fMz^3{PyDGo zfsbH9owhY4g=Fn+N*P0f4deqmO~|^vgd?8kE%h%(p}q#Wq`ICx+rxm2`7wvmR@DWhkS=}Vy{^rwY1X7tRcDpdvwdq_m%GF6@R zftg~3P8$?9D$q19r9-6z&xbc!8sgdkb=`{d98=rn)rpxp;(2SIxJbK;kpNyI{AKNd zW4tI4hvF!$Pr4S%>oHg>=efdO`#0EYN1`jMR#PF+xK$0cl6N;-B8|_}L%qrl8I83) z$V<4F@q6f@kivox_d0yy?M ze8=Dt1q5C`8u28}@ChV|UJ<#STe?)vA5r~^Ffy)vMm5^dabUi=RO51wCT3FB(d6^k~^xQWKS*C2&jfY&(z;-8~pMjSUi#h_^DG2&Gd)d zPTu)7H%I22)9tIWz}X!C8=>cusFP=P!F}{$maaU@$@~B10>tuW<<8PuzbRSHF4K3* zZCZ&DS+!5^FG?yc`3qupaF5R&Mm$La@C9C^qxxlwz5}k$SpnOns?)%)#kpn#pHFd+ zrkl;QkW14iPVbp>rLmGqQmufaEM3=b^Dpsh5SB8M3mVV->33>>9|Vf8+)vixoHO9_3Msndr%=* zY>~ZKt6+~HO)lE!aB{^XYMlV5o|8$ZQBL72Oe7-E4xQt)7%`s>$KsXXZ^xl*_~-n9EUIXNfg zAE=;VeMFjKxxt|bU$}blQJ272o0tfh#vcO$<2z;ZEM z9CSn)o&T+aX&ZF3ckkCGpq@Dsv_SN%+5$NLvJR`@jP0M{H=Rxxtr;sHZj)}NSh*VG4_JhR|imaZM+B?IoY2)lX z7K5!Ea1k9Oc7WtJ#^{pjvcQ}OA4(^d(5Oz#tg5a38sr${fei5UO4?ItiXCSt_F}T= z3I$YcNoncP-AKv_{t3@7Fu);`XX{_<@U?>n7Jv!8WyG!WRT~ubD3?hB_V^uSJaizi zNqf-znR=^ap5yb=4_BW)X$9No1ngw$Zzr0HUJs^(y1+c{4i(`uE!_de6q@7{hltqdUtw8^&B9EMH1eWGfwllVFLr^#06SEj(PkS3d#AyW^vaI`)>j6(>&pmfM~C#4syt zQvqYE;8|?W2Q~FlmUxWHsgC-69$tI2rkgo?-U=#VDnAf)_x5^Q*0MTPeGsZtaG72g zd_L=X=phN#zU+O)Gr}r=ZQA?}iOvP*5n;}$>hsUa>%ucMI9q=DdcPm?vSkWZ%k2?6 z8n!yO!{v%Mil`+$dLlxUah!U&w;>2W%W^Ebm+m{ zo|kWP7?_4i46&uiV7*mbj&Fnm!}tR1pL;XikA^X2r)NI}pTX`~)mwb>y)_|3)pC_k z!Jlq%3Om^b^7uJ7t~$Ao%H;Mmv5((d509*{;C4m;nk-0Y|6%bEgRuMD&5UoqoIg}o zvMGcflFG(?J*M+mO=bK9I#hf#+bEV=m%L~yG%?=Z!u%mHe+MS!5D)VC=QS#P zJOs6uJ7ORBAfz&3mlI%m^hS!;p?;+%^(eWEoe?PcB_1JwGNr zn#KqN*DCCc%IG)uFHFA!lcM6Pgt!+~;%j+TFp$!Cw`=OUuww6Ghv}19G^$~?ZlQze zQhl<8={M(Y?TldK-M|l^3=9ke=r~b`ai{4Y*dJ_{wVr^KF#HY%Re@?yXaV6H8o@IG zeiqKlqvNi02f#eWMw3C{GHt6wV;xt6q*KKsyNuj+3*-{kDDX3DI)7TmWmb9SXclOq z|1kB?w!wSugZ(fdV!xU<($uF=VV{a_6Qt5igqg?j8Zb# z$1ykf4d2b*OB?WHJH`A|=Y_sHM{~^7)$ZtD^~I)zW2bZVBP3CFSzS}xu8BICgjJ54 zF1f)0{6}arJw34M8wj2Zm|TA5x6a7pnwp?{@R^7U>4-pc2>mlU>x)_9&)`GjW;4!Z zdDKAInGX)IIdV@W%>AIBaO|J;f2ZJr3EMRfgcN^@g&H?%fUnH7=}h;O{{W@GU!gt8zJLyuNHh8Ha2bWp?2=}vOlUue{^Q+4i?lfjNB7$^20T_u zF99(w*?rTc{n!}s>8k_|13P8(jO`C2CQ5pWw#b&>t|sIY z+)x>>+82dq(h@WI(;)b8D=62UPYQZ=Y}T#e@#jc&-sJdtNtU9Oa^O7I{f?OkMxx}} z8LSUSkw4vTX=~cY{Pq-*!=F|GeCPbn=*cH*bKpAQpD{FEw;+W6^VAe)_dblSTlt1E zG!yqi*-5OQx9zdD2{Gdj4*B>bu@!SH>VYn*Up{|T(IeZ-)tL43=g)V6)qY-P3q2I2 z6foL_$mREyV5RJ-I)Gk_qhE@&=^Wf$m{v$WCqE07PISL%VfioeWs;U`p0o8{U5!*Y zTrV#ztxSUEY{@hOXFV_UQr~1KWSnajtT>PT#ep7Z$-hmQ^&JRwCbixtMna$LA=F=i zh`yk^e&cyUkW5~Qg&T3VAg+2}+J%7H5!U`urg&GoCYa&0@A~ zf6*Zt0kbdfEVJn^NTfR{BF{!Ar7)gT;*7)V(Jx;rkY-3rZ*}fZA$0Fe3wZf0n_h6#@$iOz91JZ_ny5`U^^GP{Nx7u zo4TQ8iDyJw`@htvTV>z)(SHI^mrJG?gn2FF=??{rtiX6{>rR53vhify-=I_zsXfWm z-x+KeA2V!%(WSuIKrySz#@||U*5@4XTRP}EV7-TR)6=S|L7S3H&E@nhz>?=<56A2} zo5au5uA3ZS5^m3H$!iglT1E*Ka9{W}JNx(3Id9^+))aeXCI@EmJ-LS~pW*A!kbMtX zHIv%VYroh*jw-cUa4RxTqQ6Piy&K$6qg-AbBL8E9k*HpKAzq9&Is64BwEeKP5qiL| zD3ebv_-@v^SVkwqX-#0nTC@^*iR+r~`o+Rr&ZL#Mk%^D{9>`IFBY;NqK{u?+X0Z zzB^^czm`hy1f^mMQ)wLG%fLX;L&C^J-9kiu>;F%*!zqHaFE8d2vs~3^v0WiA_Y(+G z#C2}+zZ_0M_sC$WdP(xXKzFCZzd<0fAv)`VBI=APWOad0U-|w)b&gn>ojP?YJ`1#S zvenv&QGNEd4u>RUp7RseFOdIDHuof{%WBZR^oPDH+DbBGDQc*J)9J8#q<-SMz2R&_ z_{GPrW_)V2C-8MVKpXi<{@3Qkyl5CUv4X;E0F0r6Y3d|9%Z0lqe!ZU+2w&wT1@eSP zafx3W+V7l=G0s>pAq!X@0d!B2Tn{n&5eJ?4*FzTRADMG?^^yoL3^&4W)xPtvr;22) zmx3U7O1`)HV$0fA7WrSzIW-aFan+?YOyy3-6!cSTK zAHovQ9>k-|wW1$Fn?>pI(TWq;Zq5o-$yfsaovQw9*wOGP9wEugdOe9B!qxk~^z|(V zA!C)4(Z1f^C3OVCff-7NL}?x+5>0od%=Ad*1z>N(XBiV_ZBRO&X!ZMW-p=GVAad3f zH+v4L2?7%;6Ra@$Fjq6s12b%Pqgc@XXP&qs7_Inr5^PNOo}31aU+bWXz2sS8sm5uWhep8GS7Px+aE)K>bV4ngjvw?3oD5LFfR!z6a zOzaDXQ^)%Q06Tlz{G0%XctyxQ{CO0G()nT^#uEJ#I!-TM?7l{xh1FJzIDvi90HM7Q zvI_TNakNBU45XK#$d0Ayqob5VEu1M9zxWVWQUl+yuR~e=JfXe(z@N`${98k<%S@gO zSD)~?bx5e$9p{Zr=Sd^9Hy{THFVKa#3Nw=EBB+44&E7pNPeRBdc*#5|z2x9ySPCmo zrW9enB^Ha+kAV;B$-m`gyEZVhvRqg`o?aiMUaXmriAAH?num}D!MLY=>h3BVx#|;l zmMRT&_cSXaPKfV!!Qs98(Z{jZ7^B6I*K&N#qNv?|4A^T9Jj`uojYu82;}Ox+J>GJ3*+eZ5f%B@?3XpMQvXk=(hB485 zUzdbg|K({si{FYe@8R7(x5$^4;5g90+FV{&S9j}ajtA%Z0?&V59EofTZGSuiUAG08 zSKfG4t?2I63q^@Os`eE2jJc`fN$k#_h^Son!}0jV{l-aI`%o3DLwGR-^gNhRGL7d! zFKWhwPrwI{ztCkLCKk;LzL1H8QP=5jr>bF;hLLd>o={idoQ|A z$z!ETsH1zvQ=SN$K7%==31eitFuhLc->el%2Zpr5KBQP1$NgAQmzTQqZpBX_`U`SQ zKCpqHXY1@Nw!T4#+n?;95|l)!{}D3k3KGqhMwh-zDXeS!Uf1{z(=>tDYj-EKIT)G) zznyv?Jk!p!2KWFZ)VD2u$spj#jlH%iwN$=4+BJmaSD_CVH!qIZGQHm4ED8@w>h@^b z4w(IHzVkr)AB^V2h>3jH+a$DiW3L^PP&p>Qq8B}5Q|c5EY(+~{A7k`Yj?QHy^7*Wi z!^NZbnHEaN^g#B(hzP0CF0H?5NeqCH43*XS$-{vAF^u=8ME#p!OMK*(RxS`Ngg9`! zSp3U{D3FHO!vRupEJ_c|au;u)|Lk|)wy~lK;*uH4{Gbjf`3=o81N4S2W>KhL4!Z>bwzci~P?>e`3u^~R=eCj3q*`<+EAPqw%~9syMvP}~yQ9qFfRf44cGz;8O| zCzI95*0h0h{1w<}Js#3=J#nXMz-dzrWyc??FdAbKZ!~!hRa*e0F~k0Dln+_wY}I=Rpj5z)G=abFwsTFvGS{4lI(% zVo-jIqIAN+&iCt$0S+e$e3Pz*k0bd_oR+8LG2mt&Z1Xz`$=9;r`= z;pdgDf3fI7gEd z;MUiycKWPj-WQxIwVlW3(~0QIU!IF7pi4QYMdVupzhRK#L#;~1*Fe+I55$P->J~GJ zPkEacMcVdj6~4E86t3TdGzY{Crakd>T$2K^%WQ|A!T>Sx94k)K>OiIKZzYOZ4FSGo z$zTS?Z~AVpuU}N6Rx~s;7+c|SW|g*_ zpB67`?b^_ARizT(;?U2>MI`br3LJu6bGqQmCaQOv{IAM?=99gT5mt72Wo4cMCU(~g zSK+MYMClaKk<+Pe6h?zy*7Ejp-=XE~6VHR=%?Hso!*8wA-?=D`ZX>kkXzmxnEWnsy z7^C)(AnmLFrW(xyJNX-ao^w`9G29^DACk0JvJXFkjXEPr-j3rltzH;4FM*cj(OP4Pj2 z=CjZ=#^euSXdohU6=B=MFQ6s}%n(Opp=Zkfedw_3iJYHb3dFcvI$T~!-hw$&C*HIM z!t6B6M832F)7b&(yxh3<8eAH+<1DC3kM-`THPpi5#ldF=NJi({;n$RHCl4pOIo>tR zZ?t?`tFGxv+LDOgJko=is^p${)C4ZrPwD7<{*VxkB9I2nZ<^j>?iBfZ$S4Pr^YiD= zxQ1dBFWh)3CAHA{32hJS&vdJ({m9JT-iY@tp?p6Wv6BT*MO$kapRbr^ci;Av-g3)Fb$6qru4H+a6v)O7v9Wq;c&SAR>;YFtGAgnS?iO8MiI|Iq_Q05Wie4?qw1^ zPm>AMjTW;R(vmp$cGi|ihKhQ@yC;)gNF{~sG@fQ9BcA~g2^A+0>1?y z^%PL(UO49s;$tdCP3?g#F=mbf{YWH^MAFWYmpK3j?Mdvl3&Cg$R0&muITh2&xjrYJ zr^p}(BUl;9iy=-z(Z(>}_qy=i(ftH1PZTQDIGNO^Yhk96c>!^`fFs|ymAJDH_0<`> zsw}JGLYHl+Mfo;Rwh6-3O@cg5NF@TGwa)0l9ueAw9 zP$Xw#_*+NWI1B%I5i4EBZe&k>ZbU5BhPQ()*tV8fD0b&A`_{a|>b;G+dnB zw>O%pt;8*cIdOC>*XM>2oinm|i!Dx+S^Bk_Cj>of&ejPSy(a9Th0=k<+U|S6*qTRt z`^pP(xK{U%4_36$8V)&DXjrFn$7 zoBifEAG*%(7T2M__$nMcqU~71DueH7zxoRGXyA`oT2%(L=yyk zZcd6ZrLvw|F}$L5zA=0~5zsaOD{ZRuqB(B+yn%+QUYno40gs+M`sj+@Vp~=6H%Ne= zfc|n$&rFS+R)>plFT40Im*ZbWB4@ggD<$c_N&+F`MWX%h5U!X5`ke{T5%V}qk`{m8 z&*ZMp*=$Z%3H=3=?Lja*fk9LhX5}cy>EJgIi2MDrlc1!=YYOSsk+Dq+KVIUzO{4{J zlp0E-2^MHU+ve0@a(FsH7=ZPfz_Vx1f{Lde$GAfe?J6m74z|7``%Mq%03OPYR!umK z{6V}Ljg4D^|2!7i?V^L+I;R(!R~EYR>p_ELi(LK!?J}L zG`xTyP&*^x?ynD_@A9Xe^ijcK$+03Us8cN31^7?|IB4)X(LzrC!MbC708zb~ojH6)No~ zQq8$@C&K)rh3k^k-RglVn~KR&U#x zq1c0@yfQ(UBq=l2{qANZ;pfP~0!>wvMq{Qx1bA5Bm7IXBd*+DVuu2noU-E_6(Q&TH z8I=V)y_(%EGEyG>$h@1X-7SH9rVL#?vY9-QZjJZmJD(hM$kP$$X{^qD9-k{(0`)#m zvj}BqSTZNf5IqEQUXzii?}WCaFc#ZvALbZ$ed37??)%^yo|>Cm{+u2}AoWmXGx#0npP+x=klq-4|2jH=;3*w_;XUzH`!raDUr>B>29KZj zFgHO`;l->>^;@90ubVr3aU?1RjjfxQ#H>vh` zuqfuCD_HBm2yg)Mj?6l46Sqc(xFR zl5lU#%o^x3#7%!8uAhCs?@2LmKzEVDZPRhB^|vS+tnMXsbDb`?sT12j0^7(5#Pu{U zY(Gv$0R2OacB_S&oTYkcdW>N5Ouda<-!538-KmZneG7C4jPU_#4dD1vn4Pq&v&!ME zX#k49V`j-&n zjL#DeS}Xa}hqpDNLvd!$;ta6YUJ>RF++QNZrIszXtkya#{E}<)zH&uR;Fz^@mFG>f3c^yt-4`rCn zbGvZk$P4BlwSpE#MA3Dh{IX-rllFf}At*8Xu7l`?bBdRl`!;8^BorJu;C^?3S527z z$q09|7I7VK{!sw)O3-vN6r`qr@V?6G`j^@&F)6-do1pM8oSBOpCCw48h%K)3e?kzya!rvbjqN0g>d?)exjv_`?gf{S&%p& zSw%-k;}p5Vennckj;8@dATS}muXs5f|d^ONqMr!VRyr{WzTGj76S zJpjm9>&#Fr;O;Nk3&bzzHjE4=@QXlaVb}PNuN59Us_{#7Pcm2H2fOS=mRL#00m-+; zLZ2T!0Bv{XW*W32(i$8awl}C+h$j$GNA!!-3lyM=DJVFbD{AL^#nAA*w45bfM%L*3 zZH>=^HN+(jl;aHz`lU4t+P^_RI^)dO{}1vpQ`FbP$!;VO_fgj7IXDKsyl5rQ&}J9A^WaWGMj@I(a47illLcFe!iXJ+ z;ezExmCUOUX^T6!>tekBo@m+ty-$@VsFc37;R-*Z*5N8c^gMBW*nE%&6A36Qn@gL& zQAwQ1O>ymWSp9;kZFjyyw{vn5-i6`Jyp7+zanOlTvQ#w_jDltk|$t!My{XP&6FPra3l zo8D(GqPo%8#DHmfUNu@g)M!44!bI*GXv;R;j9Du$+Sqf4a`+>;wDnbSKg?GsVS-fb z%BAiy{sAhA+y4~X#nnFWkYKkj;-(k87Tf?LzGcHe845t91xY1xycTLMeBZ*t8AGkJ z$L?YIO#|nZh%%cnf*c5*rV*5i0*mFzI!Jds`ua0i9?JOn2<)}InGjW*A<&JuhG^H> z=(@IQq5bP;LiW8hfhtx2y&;PF2KT1Pe*U+fD%*JWV1msNwEtn$BXy8)c;L1Liaq-8 z?ZZk5;RO$lPiC1t-`9NEyhXRoiSi4wOgGoWEMx3Ln3Xo!je&ajGzmQbzVLgS8-Q~( zEf;?dxa7LM&Z<98RV8cotF>NHM1O=6>$zMjCAee1kRL#A%k+Vb=K_q;HPvz^>zqc2c7#x^9mwl}hc37uuf;cq(t)>=rrP zm4rj3N#|97F?64GzND)+A|{BJWR+ajJFyomEsPV|Wne)MvYREu?ca-9S3xqfnl4b! zU-WJ?!X9*jcs4F_q}%TnA(fDAfGseVuovZ*ateZr*A$RBBXf^~;A!YyvTmozP+u%| zYzHIb&a=w!DItU@pFO9OLk>)R)&JydsH``(<3Pg2SRC)spbb1N zQ$7v^P0eV%XvsEs2ZWh!2)ftc_0YLU!DqE3^0yLKvetbrvR*vR*HV7M=J!E#GV`a{ z?XdQm|NV}ntlX3z`NXUl49yuvhV}VlyuTBz+f2Ei<>yPckllqt?NSVODx_n#8%)T^ z;ZLOss>pVNpVVYG`4n+I+P=*W5^+xVL*_*cd{BY6lNI0Je69)huDdZ!!d2IuldgS# zw1Uc-4FH(^3)z8|LMJ;jwF(}!VIu7JPE>bxpPsU(2BHP=&hC-dvqQe`Bn2&qO`I5U z#0ryEPJKstoCda0iW+VA*jiBXyJ}ZxFlYqW!cbJKmq_uBFPxKCaPp^LU zbp88az;*!=>{5_d`+Pegq5&T0u-AUsM2^4m(P3~NX&s=BzJ*Th-rgzfwX>ug{ei$J zuJIpUP?dSE1k6aT=nHYP zXHX~mG;Cz<0i^kSz42&{t6kgF*88=shB(hanS1Nr$`hb>{p_arYqsWn(8J?KW!2nK z8(q)AqRj*9E2wofjDN?q|Ip#Wm8AcL3we+xmfwQWTkv=^&*OU-eM3oEn$TR5e0~aW zDHY~2|Dnpu0)2=1QIGbF7r(G*lpH^Xa5J7~6u-N1x@vE0a`v9_N_(7fqPhnoak_5d zok?OsTp`au;91$cdiCi<`&J1wq>*-Le~R{>ym@Y54|6kU2Pb+ne2NmjLOSw@a7Lk~ z_4_%nfpx%QOW%|T*)jaM0fe*M+uKs*4_t?ji+rlUPLB9h>aU*IshBcAG~0$-5}_|Y z=Ij-LbD0L!KIf->+CuSH4h=Vxi>tOzkZ?wUmAZCt?sk)vt41%uLwB8Sf<{aV*6o6s znI<%TvevDERf6Adj5Wb;C91pEytwl43O2A0dg=*T6>#cF<{Y!;Tl}|f-x~MFGe`rI zqvtsh58)smTk`aIVc}&^q<@3zmA0oQc?mCAAskOX&p$LH5G*V8iy}9#6~~Djf17m_Sz=vz*t`lf`h`KC_^|$oz8RWtluhqf2YR zh}gki8chYZ{Qxp|8mk3kPImptyf&QL0ofsTZ6ZtD9_+-BAS9Vu$8Dv1BCUPJ?R}hs zG05WBDWE}?coT5;G3R-0)54=Da}NMq)l4^Km>L%snZ{4&_bnEQJT;lzsQ->F!MUEe z2f(HWVYVJJ9B1VsL5RNm%~9cFW4gy_GAy8*rY#sv^O?KOY z_3heLg!$8;XX5@`;}5G?qiqr(wT=w>G(HfarG;fa^4*8aM4@s|%@13Me>1E|QY)A!jKXJ_iYY1xQwwSDVZ9Q{)P` zn73iOlL%7mH<^qk1MI0qJJm)`l-H|+;vYPkn0V4|t3T&zW<*^^=Vhi_^w?_LQGafS z&EZK?O@2-dQG@E`GHDAFeFV%hik|s+c+A{Mo&fKFl(_c)aA5L2Njz$&Z)-?XhDcRx zhm$eGe%T(!ztxqiw}v0bit?xRSUr6z$cwRu2I7sXRX^1F@%#s6|6!G0-38u2-n9-T zI=l0e;>>`>y9so+f~xBRZGP22I<+XYdA_)rTTa=wz*a9oKQEm$b?%z#!^MzM#f)XQ z>9g_e&yy#v#kFg~o)9v6Vt64kg(2>@FVb%&6Q(ZIG-cN7ep#ASwU6%l+$W=cO zf=8&ly&YF2-Rzd_CHmas0$5j?D9f5EaV_VnmWb;?V19jq@;T0E z&VWNc*jTvDPJ1vALg^{&Lu6Ju&z3blj-I;2ZP$GG287r#QI>m(=we+O7y2+EtLc0= z^pw2ybEA3zMZ;dg<@Y`IIga+1iN>9HD*hu_((Lli#qaY$;*c^bq?ue{L}2LsXY=u7 zdn$=R7&v2Vt9q(hfB4%zVB6|LXdB2J$M`06k%k56QWC%EM72rnGtWy)eRzTHu>T4t z`mi7iKUJ{sstZ0Mge#Fi>*Z9q11h7dsWbt{Q*=4L(YI-R%_*Zq|WSRho_*3)|Ja|bUHEY zZFET;rCddcEGQ{aIo}W(m_P{ksi(?$k!sL+o3!mXs^}9VVjzl$ukpBt^X0o4M1q>*e-<%QUh8Rme!Dz*p|O~}K6cV+-X#;JsDMwb(D zH4koNOQSi6>+dBm&jh2*{%KR}!`yY|=_k(Ff-eE2sz`+lFE|Go$2Wkthv|i_jY8L{ zWSMR2v){}|J`E4QIjKV@e^O6H`Y5!NUk9cFdov0KAV&O$U_j*p>ML_I6@q${L|!am z_U2OKe^AgUrg0DKvL~=FH4r#E0=m9t$C%lvQe8d>9Yp@O>*sW`_BO=&8W=c1HE$sl zwlW=zR&sih`74mI+>bpd)ZB)B2Gf!xQ>o()|F5UF46CZ^!iDJ$=?3WrDe2xcQX*Xs z-6`EAozk63cXxM7OGzsrDIMSB{mwc5@E6zGYwtDZnq%BGmV5#DD+o<0>Z}pV{^!r1 zF2t>T`MF=HN7P6`GE%1a0j!cM)oE$S3MB;X$N|yeLitGoKCdd;l%Jcf>q=0VShGnH zQb3nL|7JG3-eP(x<>&Ba#|u+&EMiM!Vo}pVV-K7&0Qj2(m#M^bHMjWgyCLW-ljvv?_!N zxUhf*3Ltq3i#ovK!iHt{DTdICaNO%kJ=_c=cT!*_1Xna3Xs&{x_Ur*Mq{FINxvh7I$|hhYUALnhdBcRG_WsSep!%2jiATwT*2^hU(pQ=MB4zpO9h~={%0w) z!DO3v2g6g=fF))V2t!{jc&p4rAe|KPEJMv&%_5ZR5)U$e(DpZPrMmy>Ct(lu;9y06 zUpFt5M@)Du5M3=K)U>L-a(+Pcf|zFj!C+k0B|X)PxOoGWc85=8yyQAfj8#aFv@`dE zO9#;Y@KkEER1D#IIJ0+yuf44GlIsC%%Yx1WgiuSwoN*R*(;8^-JNkh7m|a2+G7Q36 z-CLy?P(#jfAArYGTNR`nBALiwOdnhxe41kif_UhbsBnDz_ZDK^4+e~ljp1u_RLZgc zUIh#oSWanoHE)q}fcPQ-*h{NhZrxr*dys~ANI~tIReJoLiM0_xbV$mdZlayustUQL zFNnmZ>bos(hkmSSNdvu$)ydz#8$jVlkZ>?`-|($Qgmf5WeLeNMuG>YO0B&U(@)>N^ zw(7Zw3BX^sg_H8&byW0o6boDOc*p?k?TecAY5pqc+_V>0!3G0q)4j=KQVa64x|3ID zmFBXqOKt(ukaN^#SPkJ-xA7&%npBbhFt*Aw;lmNop!TOd`Ra$xrYyu&zVzvfHO{kd z6?~1xxvqucghyWD4=D34IY?s8fp1|Se-(B@bF!Lf$;y@!5|^O@k?y+u z30^&Ef~KSzhxy&)V&EVlRVziT*NEoG1tM?^WmpU+eND_C*BV%!VCIi0t&1Wa`;nuf zHl|MFMuN(&mmg!o5rCT6mRvh&388hAH6b_k)s z=*`;XiDX4OcBq=n#nsBe{c~BsC;>Hf&s!v>NV5O-B_k}OfcpKvB^ei&`qMVxfYy_Q z#6>UuPjmu87t{||UhNg2TEHll-~oJ70vP@~t3#nS3Suv}NwN4XIfv@L(4!jWf{?nw z6o!&G7 zC+T-TlP6vuR&iP#Ik(9KEmfkLJej_IZtphC7^u=+rFNOHR=BTHgfx<8X^}#0K%DdqkJ7^o$#Rz$Wkjiwq1B9y-R6x5}R!5^MxZw z6WS6eBDS_o4%5D1fIss$Eew!z)Ck;KnkQQD4U4+G+}s_e zmFw?u>AR{D`Ra_pX#;`aA6(x3d|h}f-{THEU+@yfLCe~i!9Pq*-e<$Sw<2TM=ljKc%0LVcn&NuYnvT{l-h6(~H~T10F3e$W`5< z*Rb_4)6B|BtFQPFs6AR6g@6as9s7T!bE{H_cD?yiOt@c$ER9ly+eR8Pga?3a&wle% zOQ7oRYmR7V_S|7niRE+R2}$Sg<{C=+^pE&Z)7WSq$b&?AnFNPGwI&#q{-P>(9lv;Q zYVKF|ifyaAy}KkxZ;eWKC|;Q#z-2f!!Td@)tknOHo)N8T zWb^I^xeO#}ca5WOzCL zF7^*C<8cvnR_t-^30Wq4$wG06LDrBh6!+`ynS zNsd>71gmvK$6H$6EgBD46{H}V$>mi&6PFQ(g5qzE`7Nfh7F1XAlkV5Ae*fUbPDlpu zbSa1Vjq?M36poCU@^e&+!tbLry!<=!<&cT|f%|8sPdsaqAT z$v=_FTaDS%1^Mg)y4&l$tMq7tMdx%iIT#(A)bQumundL+=;Mdywn9p0eNrE*N@3%Y z!X;2QtCaFNHQsi;Fxs53OPP{PUn-Vf|Nkx(Fk|5_4$sIs;SUz7xG33rutZrGL^3D+ zKo^b=PV)&PvTOay`F|m6aU4xFau? zK7S&pVpZgy7}FK)5w`9?toyc6;97pPCeQW&2M%aI{^yDLUq$^>Ksf4$tCDbc^Br7L8b` zb_$MEC?gkOx_eW6)QSs>3WYuEFH;B^RtfpN?OSz}QsFcYwo3+wi8LY>nTlL#?75hh z#NIDGw(nzDH)i|5ep{Zj+PGR2ZGGnJ|5fGho*`%@cKFJJ-`;X;W0P@sIFv^Fh)@7K z1O@EKn_0@iF?{Nn@OUtul$M6eeWoB!xlgqUc+0SSxv1 zDuk8)&>`IVjE$!ZX_`W(?eb>eQoCAs>I7KBgqJp0+J8I^Mri^@x6CS&7LU(2z|Ra!y)d`dpdg1XY~_ay01;9RKw)Y_Pp z9|WR{SGl4_rA{LrmN0Z4YqGp6%;ChIvu^Z0c$(p|z^L5|X;9#gP;@z%I-x8;t!zY< z>Fp@d26SO^EB^xmmfhFs0~!6>j>Cq1WyqEM8#sl^M&)i`&D4!73t|1?o9*kEWSj`SNLn~CaO}q1_{>Ws2_pw;xF6{Zg-r7%ZA0wRO{-bHsqe+bkFAzT>^A~33dP>VKNJc(JLl;1TnpF#z8LF zH=%e~zNf0g!D6K^FazR?gbA%>Y_4Z55FETjkT4w)Uc1N{pxk}xdrI;Zt>u-m2i1c# zL=zZrm==ILBNytn2w+{r@RwAe?l^&}c==Oaox&mxXbd{Y=ZNrQ4SOo_d4O6@;3lv` z%|McX&iff8yAPTAPGs?4x>BvmK&sg2qi!a7H#<;zuE8eHj1BTMnMzvDi;&gcNO(8% zWI!e0QnUkuK#j|5NUe0-_LmNX(LgTKA5Q9u1 zI9@ZP6Tn-f;t02V771&%4s_cq#X>nuCqd~9fP4VI!B=JICP`4@5}v=TX3_{Q9RVI>p5MmnoXn1GxuNMxeJq1H6Et zn)jSr_tnfenIeXxh|M>&AofTmIGAKb=Ud!9ew{HcJ zQa{m%y-OG1kwk*Sh9Q)I(f)Syd|qDXO{d@K{^@4IGuI*~{d0DiK&go9lzZ308bF8I z8XN!3NSwe($;;4v9XT%q}kRKDs;MOr~jfo*DgCw32Z>;cs`V7RSaoZOmrUi z2;DOPkw1g$nhKAgctv;CcehU8*?H1Qg=Go_Ljc5w(o#YB{$FiV&H>qf(N>nf(B8@# z88mcc9(Y<1o8(04_UkJVRrdtBXsPX(fP*s_WORH#YWr9#S@?go05bSCN1hu_Lurxa zY}}~6s{=%qGIT!AcL&|fz$}k?>j}Q1!O*eCchMx>Tj%Yyri?A2q7G7Fj0RXmKFpuyFl()R$UeTs&El%Bt8oWctxF2uE zK=`AwkyC5Air|tfJ?QU@0i68YsA#SaOifL!-=kA^`24_aPJ-o_DG>Vu2ocAf?qLw{ zt6)Jq)4LiUsf+VricP3!cAk~WVrWJwLF}Z4ca1Em>d`uQ^i@ryDtI^RvxJxSwJ=S< zMs64F**V|lPVWITO&U1HYF5+lQpN9Xlb7sK-Y z>UY2OO0|SXPV;w;{YoH)jAfhqYh*G}@AR*n)MwNE6*+QU??ht z;~V2Q>tM=hcCO@ciUpgUp2VarGFi+PpVfwkd$&&;Rem=1w(Atoa4%rrdE02SmCAOH ztuEjeR92pd14%xD2*wyR9{Q&$crfQ$8I)*8AbVPdGT9f1w+sO^YZ}dH4HFJ3HaA4+ z7!uyUO(pNUPQQcXn+y6az!NmJ?;O5zs@Fi(o30VHJph8o{432i!B?}fIm%AX$pl{x zL=aXF(n5d$d*k%sxe{>SO%h7KcZOox@d)v6J$Z1!>IK5OI&is%!d0R30UTDqQZzqw z`P+5kL>*k&xC65eqBI8KJEKDD{YpKX$(}&i@18l$3X=kWE4g{aJQNhhKZGh?Pv)7O zD4@YfLl*{$FBnwdp2(`~3gCAK7z{?(RK=VNIp{9AzU+B!oKziEX&fEh(4A?v{3N^_ zmpB2`>=zDmbH$R){kJ50ROft_cbVI$@M$_@1B zRF9Am(HZ|4s2jl8BXS7O9Dd;Wz!&?26Abl{-NXXx05s<%0CYITN6oEHS>rSS#)wTm z4|9A4+lJ<5Q9SS}cxTjO^Uv+hk(3@pn^1zN1XWo{;N@|nj|UIH897xS?-oA;)n43Y z*oRxw2mkSjGD9IH|87T$dnqe;Kj%~A`oLvJa9|vY)u7%|Hs~k1K>nOyN1CBiBb?dB zH`rWuod!!|u)a(hyJWzw4F;)ucXfck*{)eUZrO<)2k3NQ?9O9Ne78%w&f=`>m$XRJ z$hGrY!k92e9~P;)QPiGEOP^!p3ccSirP$lyHJM00B+JOlLM;{q?{LzfKf_SKMg`@f zz2m+mg0DcWz*H0))gjyB*>c*#*t%QGgwM!6012AxI=Z@-Y&}kRtvSNp*Ck-Au^!y9 zVs;eHK_ZTt@HB9+dWWem{W8F`;7pfO-QyLS{<(PbB+8;)XZj@6cV^xrZs}ti*|n*} zEn+_B;uGR|+;`4_t%VsA_YXV-s~(e{6WNgmOmVhaO)xKnUzJYGWI}5z6mYyM!U`Hd z^H&AB1~&MnDt9O2sA6^JacDEnbQ1+eHv=$04n!a30HDG__`kl8Djz%W**|WcqQD)~ z?_;@jp^ygbFULSroQ+IT=wIm>LVZcd79GlOH1j!Y#_R}MO#K<4>EeK21a+qwAGtjm z{@nD?^@3L8EnI(j@XY{4?Px%av5)YQXYktTU7!)hz_rH;NX-4KYb}EXU9vN z4m~&gx=wS!Q1CXXoVV$fv$xC0F>z={@yD~!9q>L7H=m+IM*k~v3Zf!hq+ywl2*#Bt zS&{f-XQ6jIH57@)%(U}8`pA#WE!T=gPTHtFmmR(_953N&5n*K?r!ONHpUb|ieMWTp z-XCClj{(3oH0(EvA9Ssw`P;N2vB98#p|rwa|0;w>LJueEr3#=QM@fu z_D3SeK1i5p2G@y!Uk+z#t7FnBtP;0LiNoLefwag90=M;+V&=0Z!P=AC6`7aPCz|9G+kw3lf&5?3gWz(p@?2xL^GD2d9SCkw;CRU8z`YN$7gnvvB}@krQ1Vr1c7p< z8w7w^*Qz*2(OQAk#h~?=b0t*WNq&Qr*s}50bECR&6)1}Qo@QkzPC(E~xA^{;rzQ1u zvey-IrS?y#ZgH1|@8G!A6nVP(;m}^y{%Kwg7?;6o*d+TN0*(|ulL?cdv=`>$&l;V- z0pz|+cnN-b@#)JhtvGKum}PzNN+n+yy6h09@Vos52wqoVPSq(X3GLa75WH=upD^80 zbyM^`w3e4Mj%bqE4`?0)a!IwJdsN@MOh|`4nL#%NyObbPz{=y(j7|KXCXSTS=6-9H z;J>|8LvocPZ#X%I7pw$a4?agvfoua1S-kfc;Z(wtR%Q`X*3G^(C52kR5XU`3C14!U#MEtXyVZ*AoWQXp?GyTeR zl`bw-;{i(K2t-SOgX+`trFGXOT7&E9%^WFtcx2Z!Vk~?jt{x(kO#b@9Lr*^3&MWHa zYcjFsh5IW;AH8Iz&utswl&T2Q-sbh~<53D7| zboseb{6fHHe(Oo>M`dAGGPC9qJ5!S^zktSq=Hk4?*IiEy#=n8}9_Q54&p>=5Zya02 z(nrj2!4+H(5jMOYI2mtp<1NNAdc(;yz^{6$|N_e%lCem>|i36gF z%eW%$12uhzjFn1Qi$S#P_aE+s*;=y&zXvUc^gmDr)iFDem{jEgbcU7Y=@d!)JRp`tl7*&-bN#8z(223{!vADY-@_CEHqV zE>LROGqEFnX7sv=Etg!{-d5yBd@joR+Q;dTbKr_Uma%yZx0s=gr(;tEv2kC&7$5Ax z{nH@*{I=!nys^s>hBTn-&_%(@7T_-If5F#O{7gV%vcWS%sd{1GY?7-U<85%2-@?g16 zafqjnl3`Pw-tP;_ewE7aJdD^`e1+V?kbv})nf;gV9G@+?OpyiNPEY@0pf_f1MdAOV zeC+!+O>_e&)V3w|D(?mO7c=;=7Bk3ZiPtCWKdC+DXLYXxs-u8P!0!BcXzJb4PV>gBs02a^K}|!jybZ$_M)SMp!dtzjWKO z0#%dRd^&JfBHmfJ^MD_5Wap-El7;zdyjom-G3)?N1NNsU2yNuY%07d?>UX#p9C zAxJ>h7&qmV0o8c7TK@wNYt@lG&|$^&kOBj&or8lvq8ZdJ{h9wR4cuSJ1}No%M?YOE zDJ|_saB7-@ z%)f@p8d9G)|Dq?lOxZPX8sIB%#b-9GxUPQ%;3)GM{y9PVGgF43kI{|5rVZLV-edqU zyZ^OK3E6i_w5dcM8m4BL*5)djD@m%Ag?2M(rT_Z{!O?W~Tx39#Z1{H$r&**$^vb*6 zQ``boM@=ooxnTjCV3;A` zY8?jhwM;w1Gr4BgYC}B!(E8z=v3EaSSb<9oId*>ot@|(jgRCFF`KFwilqE7H*c2z@ zWgIw(<7nw0DUc@M?5j27h~sVIX?`9Qa@+L0$>RdI&(SL~z?cX9$jiBRy|ge$ z>!MQHNaKMtN#-AS^aY5?{S%sM_NC?@-WIBmoyb3BkZlQEk%=msBnL!p+pF8<(TuNq z_Rx#vk92koAvG+&2>g!ah9-b{p<7|Xc0#o*r-FXINc(G1d-cgLbZnC*f8A2tCKEG@ zV#jwF?k0wsuwdnm6?=;rtAey4{Ta6y?g%D_j{Jiaj}ORBwS?J7=z+ndTifKrmOA0l z*PkTRMwiOTN^}(6#-9=Mj1iIfS~+OUm|Hv2kfz=>U^jA~dxCq}ISqwxT*)0u7*^n= zFxj0R6^zSr3jN>7G*6&^UR_eK&1|m#C;V+1l|}}mtl%8Q2Y%T2fpvn9k@Jzc)F0I5 zN@Wt?%pky=LZuFRC`T2uJMo~j$uW)W+txi>IAxNA738&-a%?uIw~)glG$eWE1TUWO zYN_^eqIp2wlF(+n!7(RFnEC zkcqG0SrOZCnlRdHvW3O%Ae27yGz~9O!2ejVZH`u_qOMl%j7k#_6VBI#?a0M}%Uqq9 zOZn2?LAOF3@2FOV;>KNrH5+WtE${IY4GS-L2G$ymvS^kM1DQT1Pyd6w7)cP_@<)@d z>cP3=N5b3qUBwT~`FG>={H%*eG)0S*Ms7GRY#Gxe+T))bq^T+Yo%tZ4*=jg&%ve)R_iuv&MuPkg0X5!$`zMlt^vk|FEAH1979>GOv znMEGr6RB`^sWMf&x97KGw&7{Ct+^t>S+T4biChv|Shqx&4A{?9{?Ji_YFSZGlE3DH z{_Q()TL`HdKO7_+DOFmDTS-rq^fzWEj|cw7%i={te(@c(@S!!05h6-*A~3Sx4oJo0 z_k8D-7FLig>mY*e=OC>N&ARk*9D8w}iR4fH{O(xV+@i=Wrhdb`aJ!J$+s#LdrMQ+W zec@wWi7S%mX( zEIWjcEh)d$xeZyqwWxeXnKPc}N}lo(_irlS5zYvGnlW#lKZZIPfoJ3;I5?fX?5OYY zR8Dm5w*E{Pzk}${ow!)`>48jEL+n95p=nYK(V6AsG%Mf{r+iP5TQ$strazKsP=@-Y zs)=gpUfMB6nPJkQj(kF*6qW^Ff!iYAy$yR2)__jX;xp{#_>aHDTlDZ%mp5CF!zc|z z2#+w63htUSOU`%keZ;xXed=3=W(_Yqbwhh|fnA#=EU>|<_b3!CEfdeQfx%@;q_Z#; zadYzOM-(-ZXWr0f(@fP1NT2g8erzHS%kUh>k%syv$@y?cUUjnZJ#We~yJDUx?5eC~ zGRlq%lHP^Cc@!q|z&nYhDC$t>STGt61M5-#Anw=5JRjRZJ+A!TJo+Q$?Xe0je2Z)K zr90AJVdz_(vlS?f3Gw$Zd%xFWY%yjwHD@-RooX);lXykGP{rul2MhvnT`-=0xkM9X zIlph+>#fEu_WYBCt}H|#$i{v2%jo*0DDrjg5$|4LAvUhJ=MzKs+VqI^&U;w8LgLi# zg*Drv&`Z*K2cEmQC@5muxR1gxCww~LOS}9y%QQm|+ANbJ_dQW~S?Txoy?{#Og zYI!rLuO(rY@JDbPik9w35@PxuPqU^kTMt>NY{(y=TDpUtQ^pl50ua0qn0KEdjf5Np zSBxM54v5_Z1?YYvZfJh&h&imy!N}*i=jX&;L}ip$qvyW`wHN(m?G(KyrxZ~#F(Ye* zLw#YzU{-}`EUGF(D4fQug9|KIiY6@N>Rjbjc*S+;6EAIVrx@0eWzYUlVsQH@}WP1Ml95cYtCnn z*~$MCqH|~sgdv=DkNBXz;znvW`l!GooSa^iv59i$MF#|BBUuv~jX4N_UStKu? zE5dim^Wl%%O&kAAe#sUc7*p@+vW2Z6d3-tue=j+5KOn08%Qw*o`khcH+;`-KHA3zOv8V?&*{I9+j?7FRy7r9^@>=12P7mG;pIL`w z=(aEG6RJm=wPqo+hb)k3x=$7PgLBGqly;h}mof+*)Z5DfBTHAGQ3RiO1yE~I0(vc3 zt3Uf9??ntk4pe`}+Umr7XX;SKW`WbN7mu3g!&cHlZGqFmNqi@S@x zsn?13A#+1A{CtPr+0wv%5P;d)SFKdcl=Ni6yMu1aXx#1 zt`iyUOvin+r+eX=;UXfpE|8b2J!jrcL&k3KBY*6q);0`e$gMO{aa-GHFAy(zzcSj} zO%K=I9Xy}d*>2PABbU><32vyAnA_k>S`{E%V z)T>MIu4~hFRaAijP^)SZdV?y*18tVV`8m7d}LZPz9{d;^$wjDh}K z0ft<`R**gFmk11ir!15`$@!O^pW42$^(0MQ3RlQm);(lPk zTY0$_Px1>K0TfnwoP;0xmJ;caOMl3FOyV=h80a>?T6i(W1a z^7c!cr8Fqa6x0dY+iAplv*5XoBRVG~Bqn)Yd>9$AWTqj8YgFEK|Ytho#;`Rud%jc-hoMp&w8 z*Hq-CUNTTH{;UaMMa>jei{nHK0omCGhk`l)o>ewjqwLSEa<4KjCZcHan?>3vc^v|2 z+gm$>XE0+PbRo>p=H=a|QzU9lnE-i+k@bJBaX^7q(#(Q!T%p;Qjj1vQTE5v;5- zX*~payI#j)<%6(7}dpu6;(2)Uu@ru zkwa_a{PwxfjatO4x^shnsz`dD$0fC6%DusVo6T;qXk=taGx`N)b+yOPt5E?uwZO}+ z`fcg0p)8jRG?REEdoglCBZJdCL5aCfH00}nZ;$Pj@V*JlJ==?-GyKf>U@T@-F<BhJv!dXt1Vnk-mu)W!#xs=x0O#Z%fsN0O}-)KRQS8oH|GH9Rj(b1vCwnCC-b zR$s8~I)gn5fiY90D@;QD>%;*!b}40Qe%M;((@ekLyMy{mt=g(pE7^}<$G-=Jkz(t{ zPqrszD_G>%{F8|lCmwN@qn=#xQ6IRjJ~V4XFHPkYho&JwuoHp!L3a|N9E-nyN1Mvg zQ`LLU_lWJ{Nr3!}A4Qw-V`jfXy(`ow!J<%Ii9J8BHpY2X5#ka#9cd#r0%EW{isjzR zRlxcLtp(l-9>jNFJD2)kRZ{v(bw%>#uaoL-%*AezNsw8-HVqX z`PvWl9^JmnNs)J2hLD5kQ{=CiEsw*FrC5Wt4fd@;1{cSX<*kl#7m=R{w%5NH`sO~h z6Id-#ETahbJG=7}6l5#dU;T|6?c#0~jT$X0bkT;t zTsc6mI3>|_Z~inGY4tx68dfu9CdM_vnZJqn6`|U+)WR3m_nYTS;To?$=BkinQ;;uZ z7ml8LHRr>nUD4#xX?hMNiLkCxXTzjwcB{|yDYqAj#Saz6Y9BrM#hWv#I@Lb*T4TzG z_6UbUC%(wD7|s|JJGuDr*R`C4!L^-uk%@A)Otnef<7~Ky@Mp%ghgZQ-AoeF96wd_MiO0Tt zW#}xq{x?#zqMNO6`(I~{Z_{@xC@8cPIf%IW{+NKC!?XPKRgIV9I8-bDb$Q9nhP(;jz9G5;N{rVUh(E2*fY$Tqpg&*`;dlXFBk-H zqp7Y%FCvfF&xlI*+I1deoZT+qTA-TZ@lhWf*qSzKcZOWLwBo>*X%C?GR4dpDCfM57 zGyJ$yc8(?ZSobJo=`A&Io*K0CD~%YXLuq=lmp8yBxVssB^@h#nj({6O0c-3O(SG`A z%cF?a$}Ofx>TnR*Dd7uUYUj)pV6qNp5 znem_E{m11PemNn~GP0y{kD=duMVV^Ux~Kja&tJRYH)XYpg2fh;|K-^3;vp+^^-)2g z<(fccrtfPn|4qTAmPVA}FWz;e!jDAke+OaC)V3lIPRA9cY(W;Csn)CUyXZ|0bSA70A z3(PC2&XgbKcBbU4yDB1#EmP9)t5>{T0rN8tG@oB)W%G2T{R~5)ErVx2Jd44`bNyCM z*x8Fc6s&@XJs>3C?P{|;C7%!M)c=Me*?)Wp-RkNjn2&#lx%i*i+Rn;W7Q5!0xarst z-A0y9%+!~|bbDQipExQ-FAEiXOK}n?>BUL*oKHyUl7&Ii*W%6tv?F6427xw}AwC;k zx|wtZFNTcIv51AP_Mrn4{>9JmBS({;M)bUK)WWJm)uY@C(R!N;L)+f)7IZMC@}`UnnJz0+9~A?PT09|zR>(e6o7w#wi=Mq z7r&L^Qra=_vy?@cB7Buv-m3onPZhUhCS$Y-!iSEAR=BHJ4-ZSRz znU3J|lXUDzN;Fex4S0`TY{yva2Jn_qXCcmYb=~A5^`=5c^8n#U(&^!r+};Ni6eQ#ywT}eaA5F1IgNKGSy?vjNoUxe9w<)6D<9gaG zVTg5Us!lmS1#%?>q*T{<`&+!Z=p57h+_VuA3-Xyv-X8r(pgcmX!JqB226y5{enG2M ztR)#WI2W8e>rLD^TE_FAS8NnjY?u|HGw+Tli}U16{^Kpm!)L}b^2yU9eN1V}k0$wI z2^&QT*Y%_nJzm%dNIuNB&64k7?*3v=bratyNZ)(qMjuzI`U&}_t3b8B1D*3bi@a0@ zC1^L?)lV$6F9Y~=98;`VZuqMoq*NmU-s9d8D6;O5MhzzJ!6b>~;j+p#K4u@{D)cAN z`7mln^>`%JG`z`wS@|Nhx@Yo<+dW68xAiGhA>Ln`3$^AY^l{5xnoYc0!sOeG7aVa| zR6?9Sjbyy<6Tv6<`a8blXXdPwF~_2^7_L57L++l!$1Wx0*f5+;K(l;nP8Up-hij-$ zaO24p1>@rwrz~j*vkeX&&?Ngb)nVrCW*%~NFHWUV4#v@*roDE*GX+P@Ag~*9sz+(z zATthzq)VGI*YzU&$5ewlt3|52yEe|K+U`X~9R9>8Z^znjdtGJhpq=nhmC%U?etw+t zN;DJOsRR8)(E;K|I~`wHi7(X(v#Qllvpi3Xui&aEb!x+y49`TXC_;?K&5fO|dn$?g z+p6aUx7Ga6-%T^Yt=mW{-%&_@(dMHK?g7=+I?b;mJ@l}eZGlGUi*_O2QzGG?ti=PS zXI8sDH#EsIW!xmqZ(ju4+ZmNmBU?sOgTImcAcxWu02cyAhN*Hv13|kU^oi1Lct#Oy z+29-jVYoB#hR(3Jq8Jf!fi-hbS@v2a)qd>v77dmwQRbrcJ5C)z?@K888}*XicqugT zgtTtuHfpmv_Bj9uP}E5#h_1vjvdIWzbSM%f(>6)``PI+-+v@B22PS63h1?Yk{=J*2 zlr@$l>bwsatABNt$iWLFY+df#fErK8vxVv>G@ba-26w1lO3Nx1`hSxC78F+#8Fo|a zIQ;j9F)&KoQlhPW;*#|kn;?PBjg6|KV9oEH#i+&Y|* zjvl}1EtQUikf+~&gzMH~uc6jsB8J$+;6;lDv0ky_mYJWlF6m~c&L5YV&96qx{fANO zApUIYFvC0CJ(x%%8Tgw4Q@gY=dB@ykH}&yd2K$D%l|-LTmS({~Jq6;h%cOH~eCwsu zHfzF?CLzKjYOBt_!Geq6+0{FtPth+b!Dklt8Xi$0Fes-ruoT&~*8!hpQnEstRn0=4c!?=gRUaq61dpk(eQQ`Pq~#z&=2q2C&y4sQ54%14 z`ZrkkP|sF#WTnue(mRD&Fk|3SU=G_xt8?85FB3stqjPtW;9S@(G%>|w1_kSYxRq9l zYFOSt|7=ft>^_fsPF&Za7rp!|t}`@TZRz7;A<~!nom~+QU!mTK)9Z$K~*S^M(+!Ad$g+fv8}z1efD-hKY_@ZA>X0Uzp<*M?;pB=P7aqz5`FAb<=h=?4v@< zj<^6a@WK8@oYURH`+ErOan3%o%=(8Y&;R56^j-Ejs*%n{8Y!`3BfCoo0$qB!m4 zRy1=AoX+H432cZ%10F9t@s791YEik;F8q|bEKljs43VeQ+{1>Dj>rJ^yij+ncLaQa z`;(?K1lq`aB(3F-BzAo}>TuCg#h3)Jr_jT5dFf~f1zEHX)c-c98FJ@m1lE{iiViYo zqb{N6a$*=9V`%{*%t0cxGwW0A~=dI+F^ z#wpcnd&G7Wa-?7V;7V;6&|k~0#KrOE7UK*m!;&B`!;K!AM0MrXG5+KQf;;!<g>?1!&?yE7X?gw@7!C*#V>@%%g=V5@s(-(jpGId5uSYTR$zQ>bMal15fjH zey*F2e|nA|^WFPcay8FrsmKBeS}`oieM6`q(X*gRGc{GhGE*koc7o(5FpGMP%;|47 zqhtj)=N2uAvA;7LYzQS3L;jpy5>Q#}sr~i%b{YCE!dH6w{G8){LsUvBzkha_P<;BF z1Fp0smr77%f8fRtz;Yb*@5h9jP-}!U2=HOKMyDuHXFARpbuPcaG8eYexjz$z;9suY znb`Zh;3<$1D!BQsko%rS%`B%y3=Pwp4$Gj7QgjXv-^hH z+Xg-N@2GxOVu(gZUG<5?-wW;0H8pHnigz)0pLl8-kZg%M5f;&)6ci!l7I}^tk|6-n z4)i_4`kj8`-~L5`A7)7t1^v-vY$o0{^giGs+o2xv`*%bxrYWwGH4jLj)(|C%I}Z=P v!}Z}r^S$|Xdkfxc`R{I&arN+j-roANX1B>E{~aKQ0{`Tslps|SMnV4{a`>6R literal 0 HcmV?d00001 diff --git a/images/neural-networks.svg b/images/neural-networks.svg new file mode 100644 index 0000000..9c3cb3c --- /dev/null +++ b/images/neural-networks.svg @@ -0,0 +1,10131 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + image/svg+xml + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/notebooks_en/Appendix.ipynb b/notebooks_en/Appendix.ipynb new file mode 100644 index 0000000..82b67d8 --- /dev/null +++ b/notebooks_en/Appendix.ipynb @@ -0,0 +1,155 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "cc66e939", + "metadata": {}, + "source": [ + "## Beyond the math: all the fancy terms" + ] + }, + { + "cell_type": "markdown", + "id": "ecaec976", + "metadata": {}, + "source": [ + "Though knowing math is enough for us to implement and use neural network models, understanding the jargon helps communicate with people from different backgrounds. In this section, we would like to fill in the gap between math and the terminology commonly used in the community of machine learning and deep learning.\n", + "\n", + "Let's use the math from a fully connected neural network as an example:\n", + "\n", + "$$\n", + "\\begin{aligned}\n", + "\\mathbf{z}^1 &= \\sigma_0\\left(\\left(W^0\\right)^\\mathsf{T}\\mathbf{x} + \\mathbf{b}^0\\right) \\\\\n", + "\\mathbf{z}^2 &= \\sigma_1\\left(\\left(W^1\\right)^\\mathsf{T}\\mathbf{z}^1 + \\mathbf{b}^1\\right) \\\\\n", + "&\\vdots \\\\\n", + "\\hat{\\mathbf{y}} &= \\sigma_L\\left(\\left(W^L\\right)^\\mathsf{T}\\mathbf{z}^L + \\mathbf{b}^L\\right)\n", + "\\end{aligned}\n", + "$$\n", + "\n", + "Often you see people trying to use this kind of graph to explain a fully connected neural network:\n", + "\n", + " \n", + "\n", + "This graphical illustration may come from the fact that neural network models are inspired by the true neural networks in human bodies. However, this figure does not really tell us how exactly a model should be implemented.\n", + "\n", + "Nevertheless, the graphical illustration makes it's easier to understand why the deep learning community names the math components with the following terms." + ] + }, + { + "cell_type": "markdown", + "id": "13d31e91", + "metadata": {}, + "source": [ + "##### Neurons\n", + "\n", + "Each value in the intermediate and final results is called a neuron. Though we use vectors $\\mathbf{z}^1$, $\\dots$, $\\hat{\\mathbf{y}}$ to denote the calculation results, they are composed of many elements. In other words, elements in vectors are called neurons. For example, if $\\mathbf{z}^1$ has $12$ elements, i.e., $n_1=12$, then we say $\\mathbf{z}^1$ has 12 neurons.\n", + "\n", + "A neuron in a neural network means it is a signal processing node. It receives values/signals from other neurons, does some calculations, and then provides the calculation result (or says processed signal) to other neurons. For example, in the above fully connected neural network model, if we expand the vector-matrix form to explicit equations, the element $z_1^2$ from the intermediate vector $\\mathbf{z}^2$ is obtained through:\n", + "\n", + "$$\n", + "z_1^2 = \\sigma_0\\left(W_{1,1}^{1}z_1^{1}+W_{2,1}^{1}z_2^1+\\cdots+W_{n_1,1}^1 z_{n_1}^1 + b_1^1\\right)\n", + "$$\n", + "\n", + "So we say the neuron $z_1^2$ receives signals from neurons $z_1^1$, $z_2^1$, $\\cdots$, $z_{n_1}^1$. And because the value of neuron $z_1^2$ is needed to calculate the intermediate vector $\\mathbf{z}^3$, we also say the neuron $z_1^2$ provides the signal to neurons $z_1^3$, $z_2^3$, $\\cdots$, $z_{n_3}^3$ for further signal processing." + ] + }, + { + "cell_type": "markdown", + "id": "ce6812fe", + "metadata": {}, + "source": [ + "##### Layers\n", + "\n", + "Neurons that are independent of each other are put together and becomes a layer. For example, to calculate $z_1^2$ from the input $\\mathbf{x}$, we don't need the value of $z_2^2$, and vice versa. We say $z_1^2$ and $z_2^2$ belong to the same layer. In fact, if we look at the vector-matrix format of the model, we can see that each vector is a layer, e.g., $\\mathbf{z}^1$ is a layer because all elements in the vector $\\mathbf{z}^1$ are independent of each other. The same applies to $\\mathbf{z}^2$, $\\dots$, $\\hat{\\mathbf{y}}$.\n", + "\n", + "We call the vector $\\hat{\\mathbf{y}}$ the output layer because it's the final output of a neural network. Some people extend the naming system to the input vector $\\mathbf{x}$ and call it an input layer. " + ] + }, + { + "cell_type": "markdown", + "id": "860d9943", + "metadata": {}, + "source": [ + "##### Hidden layers\n", + "\n", + "Intermediate vectors $\\mathbf{z}^1$, $\\dots$, $\\mathbf{z}^L$ are called hidden layers because we don't see them if we treat a neural network as a black box. The variable $L$ hence denotes the number of hidden layers." + ] + }, + { + "cell_type": "markdown", + "id": "cf50d014", + "metadata": {}, + "source": [ + "##### Forward propagation\n", + "\n", + "Forward propagation means the procedure of the output $\\hat{\\mathbf{y}}$, starting from the input $\\mathbf{x}$, then $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, and so on. The word propagation may come from the propagation of signals. Each neuron layer receives signals from the previous layer, does some processing, and then passes the processed signal to the next layer." + ] + }, + { + "cell_type": "markdown", + "id": "241403b9", + "metadata": {}, + "source": [ + "##### Backward propagation\n", + "\n", + "Backward propagation is related to calculating the gradients of parameters. We didn't see it in our teaching material as we rely on `autograd` to do the job. However, in a nutshell, backward propagation is a technique for obtaining the gradients of parameters. It is an application of the chain rule from calculus. Nowadays, it's more common to use third-party libraries for calculating gradients, and the backward propagation is usually how these libraries work under the hood." + ] + }, + { + "cell_type": "markdown", + "id": "e3032d86", + "metadata": {}, + "source": [ + "##### Training and learning\n", + "\n", + "You may notice we use the term *optimization* most of the time, though sometimes the terms *training* and *learning* slip through and are present in the teaching material. They all mean the same thing in machine learning: finding the best set of parameters that makes a model best fits given data. In other words, if you have taken any courses in numerical methods or numerical analysis, they are synonyms of *model fitting*.\n", + "\n", + "In lesson 5, we used the example of take-home exercises, quizzes, and final exams to explain the concepts of training, validation, and test datasets. We can see the optimization of a model is indeed similar to training a student to some degree." + ] + }, + { + "cell_type": "markdown", + "id": "dbdcb6fa", + "metadata": {}, + "source": [ + "##### Hyperparameters\n", + "\n", + "Matrices and vectors, $W^0$, $\\mathbf{b}^0$, $W^1$, $\\mathbf{b}^1$, etc., are called *model parameters* or simply *parameters*. However, we also have other parameters such as the coefficient used in gradient descent (i.e., learning rate), the coefficients for regularization, etc. These parameters are not part of a model and are called hyperparameters." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e97a8795", + "metadata": {}, + "outputs": [], + "source": [ + "# Execute this cell to load the notebook's style sheet, then ignore it\\n\",\n", + "from IPython.core.display import HTML\n", + "css_file = '../style/custom.css'\n", + "HTML(open(css_file, \"r\").read())" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.5" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} From ba2ac6940074432af1dbd9027b5f0098497fe887 Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Fri, 6 Aug 2021 02:42:28 -0400 Subject: [PATCH 3/8] tune parameters and content - make lesson 5 and lesson 7 more comparabnle regarding the optimization - use n1, n2, ... to replace nz1, nz2, ... - monitor non-regularized loss - change the loss to mean, instead of summation - also use mean for the regularization term - change random seed --- .../5_Multiple_Logistic_Regression.ipynb | 161 +++++--- .../7_The_First_Deep_Neural_Network.ipynb | 368 +++++++++++++++--- 2 files changed, 425 insertions(+), 104 deletions(-) diff --git a/notebooks_en/5_Multiple_Logistic_Regression.ipynb b/notebooks_en/5_Multiple_Logistic_Regression.ipynb index 10144ad..148ffbc 100644 --- a/notebooks_en/5_Multiple_Logistic_Regression.ipynb +++ b/notebooks_en/5_Multiple_Logistic_Regression.ipynb @@ -183,7 +183,7 @@ "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -219,7 +219,7 @@ "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -360,19 +360,18 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The loss function is also the same as in lesson 2 for regular logistic regression, but for multiple features, plus we add the regularization term from lesson 4:\n", - "\n", + "The loss function is also the same as in lesson 2:\n", "$$\n", - "\\mathrm{loss} = - \\sum_{i=1}^{N} y_{\\text{true}}^{(i)}\\log\\left(\\hat{y}^{(i)}\\right) + \\left(1-y_{\\text{true}}^{(i)}\\right)\\log\\left(1-\\hat{y}^{(i)}\\right) + \\lambda\\sum_{i=1}^{n}{w_i}^2\n", + "L = - \\frac{1}{N} \\sum_{i=1}^{N} y_{\\text{true}}^{(i)}\\log\\left(\\hat{y}^{(i)}\\right) + \\left(1-y_{\\text{true}}^{(i)}\\right)\\log\\left(1-\\hat{y}^{(i)}\\right)\n", "$$\n", "\n", "Or, in vector form:\n", "\n", "$$\n", - "\\mathrm{loss} = - \\left[\n", + "L = - \\frac{1}{N} \\left[\n", "\\mathbf{y}_{\\text{true}}\\cdot\\log\\left(\\mathbf{\\hat{y}}\\right)+\n", "\\left(\\mathbf{1}-\\mathbf{y}_{\\text{true}}\\right)\\cdot\\log\\left(\\mathbf{1}-\\mathbf{\\hat{y}}\\right)\\\n", - "\\right] + \\lambda\\sum_{i=1}^{n}{w_i}^2\n", + "\\right]\n", "$$\n", "\n", "where the bolded $\\mathbf{1}$ represents a vector of ones." @@ -384,7 +383,7 @@ "metadata": {}, "outputs": [], "source": [ - "def model_loss(x, true_labels, params, _lambda=1.0):\n", + "def model_loss(x, true_labels, params):\n", " \"\"\"Calculate the predictions and the loss w.r.t. the true values.\n", " \n", " Arguments\n", @@ -397,8 +396,6 @@ " params : a tuple/list of two elements\n", " The first element is a 1D array with shape (n_total_pixels). The\n", " second elenment is a scalar.\n", - " _lambda : float\n", - " The weight of the regularization term. Default: 1.0\n", " \n", " Returns\n", " -------\n", @@ -407,14 +404,59 @@ " \"\"\"\n", " pred = logistic_model(x, params)\n", " \n", + " n_images = x.shape[0]\n", + " \n", " loss = - (\n", " numpy.dot(true_labels, numpy.log(pred+1e-15)) +\n", " numpy.dot(1.-true_labels, numpy.log(1.-pred+1e-15))\n", - " ) + _lambda * numpy.sum(params[0]**2)\n", + " ) / n_images\n", " \n", " return loss" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "To avoid overfitting, we add regularization to the loss, which was introduced in lesson 4:\n", + "\n", + "$$\n", + "L^{reg} = L + \\frac{\\lambda}{n}\\sum_{i=1}^{n}{w_i}^2\n", + "$$" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [], + "source": [ + "def regularized_loss(x, true_labels, params, _lambda=1.):\n", + " \"\"\"Return the loss with regularization.\n", + " \n", + " Arguments\n", + " ---------\n", + " x, true_labels, params :\n", + " Parameters for function `model_loss`.\n", + " _lambda : float\n", + " The weight of the regularization term. Default: 1.0\n", + " \n", + " Returns\n", + " -------\n", + " loss : a scalar\n", + " The summed loss.\n", + " \"\"\"\n", + " loss = model_loss(x, true_labels, params) \n", + " return loss + _lambda * numpy.mean(params[0]**2)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Compared to the equation in lesson 4, we have an extra $\\frac{1}{n}$ in the weight of the regularization term. We do this to keep $\\lambda=1$ but still have the magnitude of the regularization term comparable to the original loss $L$. $\\frac{1}{n}$ can be absorbed into $\\lambda$ if you want -- $\\lambda$ is a hyperparameter that you can tune and freely choose the value." + ] + }, { "cell_type": "markdown", "metadata": {}, @@ -422,7 +464,8 @@ "##### Notes:\n", "\n", "1. We added a tiny term `1e-15` to the `log` calculation to avoid the infinity when `pred` or/and `1-pred` are zero.\n", - "2. The function `model_loss` combines the calculations of model predictions and the loss. By doing so, we can easily calculate the gradients using the `grad` from `autograd`, just like we did in lesson 2." + "2. Why don't we combine `model_loss` and `regularized_loss`? When tracking the history of losses during optimization, we may need original losses (i.e., non-regularized), so we need a function to give us non-regularized losses.\n", + "3. The functions `model_loss` and `regularized_loss` combine the calculations of model predictions and the loss. By doing so, we can easily calculate the gradients using the `grad` from `autograd`, just like we did in lesson 2." ] }, { @@ -451,7 +494,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 13, "metadata": {}, "outputs": [ { @@ -496,7 +539,7 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 14, "metadata": {}, "outputs": [], "source": [ @@ -513,7 +556,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 15, "metadata": {}, "outputs": [], "source": [ @@ -546,7 +589,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 16, "metadata": {}, "outputs": [], "source": [ @@ -584,7 +627,7 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 17, "metadata": {}, "outputs": [], "source": [ @@ -610,7 +653,7 @@ }, { "cell_type": "code", - "execution_count": 17, + "execution_count": 18, "metadata": {}, "outputs": [], "source": [ @@ -697,7 +740,7 @@ }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 19, "metadata": {}, "outputs": [], "source": [ @@ -762,12 +805,12 @@ }, { "cell_type": "code", - "execution_count": 19, + "execution_count": 20, "metadata": {}, "outputs": [], "source": [ "# a function to get the gradients of a logistic model\n", - "gradients = grad(model_loss, argnum=2)\n", + "gradients = grad(regularized_loss, argnum=2)\n", "\n", "# initialize parameters\n", "w = numpy.zeros(images_train.shape[1], dtype=float)\n", @@ -783,7 +826,7 @@ }, { "cell_type": "code", - "execution_count": 20, + "execution_count": 21, "metadata": {}, "outputs": [ { @@ -834,14 +877,16 @@ "source": [ "## Training/optimization\n", "\n", - "In the optimization loop (using gradient descent), we'll monitor the progress of training using the validation dataset. If the validation loss is not changing much, we'll stop the training.\n", + "In the optimization loop (using gradient descent), we'll monitor the progress of training using the validation dataset. If the validation loss is not changing much, we'll stop the training. The loss we want to monitor is the original loss. On the other hand, when calculating the gradients, we use the regularized loss against the training data.\n", + "\n", + "Here, we only use validation loss to control the optimization. In real applications, it's common to use a combination of validation loss, accuracy, and other metrics to control the optimization.\n", "\n", - "Here, we only use validation loss to control the optimization. In real applications, it's common to use a combination of validation loss, accuracy, and other metrics to control the optimization." + "Also, the step size (the variable `lr`) is `1e-2`. The step size affects whether we can and how fast we'll find the solution. Finding a proper step size requires hyperparameter tuning mentioned previously. Here we just give you a proper value we found as hyperparameter tuning is out of the scope of this lesson." ] }, { "cell_type": "code", - "execution_count": 21, + "execution_count": 22, "metadata": { "scrolled": true }, @@ -850,21 +895,16 @@ "name": "stdout", "output_type": "stream", "text": [ - "10...20...30...40...50...60...70...80...90...100...110...120...130...140...150...160...170...180...190...200...210...220...230...240...250...260...270...280...290...300...310...320...330...340...350...360...370...380...390...400...410...420...430...440...450...460...470...480...490...500...510...520...530...540...\n", - "\n", - "Upon optimization stopped:\n", - " Iterations: 541\n", - " Validation loss: 126.7753373500034\n", - " Validation precision: 0.900709219858156\n", - " Validation recall: 0.8141025641025641\n", - " Validation F-score: 0.8552188552188552\n", - " Change in validation loss: 2.0973401770177472e-06\n" + "10...20...30...40...50...60...70...80...90...100...110...120...130...140...150...160...170...180...190...200...210...220...230...240...250...260...270...280...290...300...310...320...330...340...350...360...370...380...390...400...410...420...430...440...450...460...470...480...490...500...510...520...530...540...550...560...570...580...590...600...610...620...630...640...650...660...CPU times: user 34.2 s, sys: 412 ms, total: 34.6 s\n", + "Wall time: 5.77 s\n" ] } ], "source": [ - "# learning rate\n", - "lr = 1e-5\n", + "%%time\n", + "\n", + "# step size\n", + "lr = 1e-2\n", "\n", "# a variable for the change in validation loss\n", "change = numpy.inf\n", @@ -878,7 +918,7 @@ "# keep running if:\n", "# 1. we still see significant changes in validation loss\n", "# 2. iteration counter < 10000\n", - "while change >= 1e-5 and i < 10000:\n", + "while change >= 1e-6 and i < 10000:\n", " \n", " # calculate gradients and use gradient descents\n", " grads = gradients(images_train, labels_train, (w, b))\n", @@ -901,13 +941,40 @@ " \n", " # print the progress every 10 steps\n", " if i % 10 == 0:\n", - " print(\"{}...\".format(i), end=\"\")\n", + " print(\"{}...\".format(i), end=\"\")" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "\n", + "Upon optimization stopped:\n", + " Iterations: 666\n", + " Validation loss w/o regularization: 0.5188273788049017\n", + " Validation loss w/ regularization: 0.04127987992406643\n", + " Validation precision: 0.8951048951048951\n", + " Validation recall: 0.8205128205128205\n", + " Validation F-score: 0.8561872909698997\n", + " Change in validation loss: 9.484651848988673e-07\n" + ] + } + ], + "source": [ + "val_loss_reg = regularized_loss(images_train, labels_train, (w, b))\n", "\n", "print(\"\")\n", "print(\"\")\n", "print(\"Upon optimization stopped:\")\n", "print(\" Iterations:\", i)\n", - "print(\" Validation loss:\", val_loss)\n", + "print(\" Validation loss w/o regularization:\", val_loss)\n", + "print(\" Validation loss w/ regularization:\", val_loss_reg)\n", "print(\" Validation precision:\", score[0])\n", "print(\" Validation recall:\", score[1])\n", "print(\" Validation F-score:\", score[2])\n", @@ -930,16 +997,16 @@ }, { "cell_type": "code", - "execution_count": 22, + "execution_count": 24, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Final precision: 88.0%\n", + "Final precision: 88.6%\n", "Final recall: 84.6%\n", - "Final F-score: 86.3%\n" + "Final F-score: 86.6%\n" ] } ], @@ -957,7 +1024,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Bravo! The precision improved from $60.2\\%$ to $88.0\\%$, which means when our model says a part has defects, there is an $88.0\\%$ chance it is indeed defective. The recall dropped from $100%$ to $84.6\\%$, which means our model misses about $15\\%$ of the defective parts. The F-score, representing overall accuracy, improved from $75.2\\%$ to $86.3\\%$." + "Bravo! The precision improved from $60.2\\%$ to $88.6\\%$, which means when our model says a part has defects, there is an $88.6\\%$ chance it is indeed defective. The recall dropped from $100%$ to $84.6\\%$, which means our model misses about $15\\%$ of the defective parts. The F-score, representing overall accuracy, improved from $75.2\\%$ to $86.6\\%$." ] }, { @@ -966,7 +1033,7 @@ "source": [ "##### Final note\n", "\n", - "No model is perfect and it's not realistic to expect $100\\%$ accuracy. How accurate a model should be to consider it a good model is problem-dependent. Still, an F-score of $86.3\\%$ does not seem very exciting. In this example, it means about $15\\%$ of defective products slip through and may be handed over to customers.\n", + "No model is perfect, and it's not realistic to expect $100\\%$ accuracy. How accurate a model should be to consider it a good model is problem-dependent. Still, an F-score of $86.6\\%$ does not seem very exciting. And it's not pleasant to see about $15\\%$ of defective products slip through and may be handed over to customers. \n", "\n", "In a later lesson, we will improve the performance by replacing the multiple logistic regression model with something more interesting: a **neural network** model." ] @@ -996,7 +1063,7 @@ }, { "cell_type": "code", - "execution_count": 23, + "execution_count": 25, "metadata": {}, "outputs": [ { @@ -1156,7 +1223,7 @@ "" ] }, - "execution_count": 23, + "execution_count": 25, "metadata": {}, "output_type": "execute_result" } @@ -1185,7 +1252,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.9.5" + "version": "3.9.6" } }, "nbformat": 4, diff --git a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb index 7a42e46..835bdc0 100644 --- a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb +++ b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb @@ -30,7 +30,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 1, "metadata": {}, "outputs": [], "source": [ @@ -48,11 +48,11 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 2, "metadata": {}, "outputs": [], "source": [ - "numpy.random.seed(0)" + "numpy.random.seed(10000)" ] }, { @@ -64,7 +64,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 3, "metadata": {}, "outputs": [], "source": [ @@ -89,7 +89,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 4, "metadata": {}, "outputs": [], "source": [ @@ -214,9 +214,9 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "- Lengths of intermediate vectors: $n_{z^1}$, $n_{z^2}$, $\\dots$, $n_{z^L}$\n", + "- Lengths of intermediate vectors: $n_1$, $n_2$, $\\dots$, $n_L$\n", " \n", - " $n_{z^1}$, $n_{z^2}$, $\\dots$, $n_{z^L}$ are not visible in the above matrix-vector form of the model. They represent the length of the intermediate vectors $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. These variables are user-defined and also control the complexity of models." + " $n_1$, $n_2$, $\\dots$, $n_L$ are not visible in the above matrix-vector form of the model. They represent the length of the intermediate vectors $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. These variables are user-defined and also control the complexity of models." ] }, { @@ -241,7 +241,7 @@ "\n", "1. $L=1$,\n", "2. $\\sigma_0(\\mathbf{g})=\\sigma_1(\\mathbf{g})=\\operatorname{logistic}(\\mathbf{g})$, and\n", - "3. $n_{z^1}=64$, meaning $\\mathbf{z}^1=\\begin{bmatrix}z_1^1 & z_2^1 & \\cdots & z_{64}^1 \\end{bmatrix}^\\mathsf{T}$\n", + "3. $n_1=64$, meaning $\\mathbf{z}^1=\\begin{bmatrix}z_1^1 & z_2^1 & \\cdots & z_{64}^1 \\end{bmatrix}^\\mathsf{T}$\n", "\n", "Also, each of our images has a total of $128\\times 128=16384$ pixels (revisit lesson 5 if you don't remember), so $\\mathbf{x}=\\begin{bmatrix}x_1 & x_2 & \\cdots & x_{16384}\\end{bmatrix}^\\mathsf{T}$.\n", "\n", @@ -369,7 +369,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 5, "metadata": {}, "outputs": [], "source": [ @@ -406,11 +406,11 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 6, "metadata": {}, "outputs": [], "source": [ - "def model_loss(x, true_labels, params, alpha=1.0):\n", + "def model_loss(x, true_labels, params):\n", " \"\"\"Calculate the predictions and the loss w.r.t. the true values.\n", " \n", " Arguments\n", @@ -421,10 +421,10 @@ " The true labels of the input images. Should be 1D and have length of\n", " n_images.\n", " params : a tuple/list of two elements\n", - " The first element is a 2D array with shape (n_total_pixels, 1). The second\n", - " elenment is a scalar.\n", - " alpha : float\n", - " The weight of the regularization term. Default: 1.0\n", + " - The first element is W0, a 2D array with shape (n_total_pixels, n_z1).\n", + " - The second elenment is b0, an 1D array with length n_z1.\n", + " - The third element is W1, an 1D array with length n_z1.\n", + " - The fourth element is b1, a scalar.\n", " \n", " Returns\n", " -------\n", @@ -433,16 +433,42 @@ " \"\"\"\n", " pred = neural_network_model(x, params)\n", " \n", + " n_images = x.shape[0]\n", + " \n", " # major loss\n", " loss = - (\n", " numpy.dot(true_labels, numpy.log(pred+1e-15)) +\n", " numpy.dot(1.-true_labels, numpy.log(1.-pred+1e-15))\n", - " )\n", - "\n", - " # regularization\n", - " reg = (params[0]**2).sum() + (params[2]**2).sum()\n", + " ) / n_images\n", + " \n", + " return loss" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [], + "source": [ + "def regularized_loss(x, true_labels, params, _lambda=1.):\n", + " \"\"\"Return the loss with regularization.\n", + " \n", + " Arguments\n", + " ---------\n", + " x, true_labels, params :\n", + " Parameters for function `model_loss`.\n", + " _lambda : float\n", + " The weight of the regularization term. Default: 0.01\n", " \n", - " return loss + alpha * reg" + " Returns\n", + " -------\n", + " loss : a scalar\n", + " The summed loss.\n", + " \"\"\"\n", + " loss = model_loss(x, true_labels, params)\n", + " Nw = params[0].shape[0] * params[0].shape[1] + params[2].size\n", + " reg = ((params[0]**2).sum() + (params[2]**2).sum()) / Nw\n", + " return loss + _lambda * reg" ] }, { @@ -461,30 +487,30 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "# a function to get the gradients of a logistic model\n", - "gradients = grad(model_loss, argnum=2)" + "gradients = grad(regularized_loss, argnum=2)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "For our convenience, we use variables `nz0` to store the length of input vector $\\mathbf{x}$ (recall that $\\mathbf{z}^0\\equiv\\mathbf{x}$) and use `nz1` to represent $n_{z^1}$:" + "For our convenience, we use variables `n0` to store the length of input vector $\\mathbf{x}$ (recall that $\\mathbf{z}^0\\equiv\\mathbf{x}$) and use `n1` to represent $n_1$:" ] }, { "cell_type": "code", - "execution_count": null, + "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "# number of elements in z0 (i.e., x), z1, ...\n", - "nz0 = images_train.shape[1] # i.e., res * res\n", - "nz1 = 64" + "n0 = images_train.shape[1] # i.e., res * res\n", + "n1 = 64" ] }, { @@ -503,14 +529,14 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "# initialize parameters\n", - "W0 = numpy.zeros((nz0, nz1))\n", - "b0 = numpy.zeros(nz1)\n", - "W1 = numpy.zeros(nz1)\n", + "W0 = numpy.zeros((n0, n1))\n", + "b0 = numpy.zeros(n1)\n", + "W1 = numpy.zeros(n1)\n", "b1 = 0.0" ] }, @@ -523,9 +549,18 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 11, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Gradients of W0 are zeros: True\n", + "Gradients of W1 are zeros: False\n" + ] + } + ], "source": [ "grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", "print(\"Gradients of W0 are zeros:\", numpy.allclose(grads[0], 0.))\n", @@ -539,7 +574,7 @@ "And how does this issue matter? Recall how we perform gradient descent -- we update parameters by substracting gradients from the current values of parameters:\n", "\n", "$$\n", - "\\text{new }W^0 = \\text{current }W^0 - \\text{learning rate}\\times\\text{gradients of }W^0\n", + "\\text{new }W^0 = \\text{current }W^0 - \\text{step size}\\times\\text{gradients of }W^0\n", "$$\n", "\n", "So if the gradients of $W^0$ are zeros, then gradient descent will not change the values of $W^0$ at all. In other words, our model will not improve no matter how many iterations we run for optimization." @@ -564,8 +599,7 @@ "metadata": {}, "source": [ "$$\n", - "W^i = \\text{random numbers uniformly drawn from} \\\\\n", - "\\text{interval }\\left[-\\sqrt{\\frac{6}{n_{z^i}+n_{z^{i+1}}}}, \\sqrt{\\frac{6}{n_{z^i}+n_{z^{i+1}}}}\\right]\n", + "W^i = \\text{random numbers uniformly drawn from interval }\\left[-\\sqrt{\\frac{6}{n_i+n_{i+1}}}, \\sqrt{\\frac{6}{n_i+n_{i+1}}}\\right]\n", "$$\n", "and\n", "$$\n", @@ -576,17 +610,17 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "# initialize parameters\n", - "scale = numpy.sqrt(6/(nz0+nz1))\n", - "W0 = numpy.random.uniform(-scale, scale, (nz0, nz1))\n", - "b0 = numpy.zeros(nz1)\n", + "scale = numpy.sqrt(6/(n0+n1))\n", + "W0 = numpy.random.uniform(-scale, scale, (n0, n1))\n", + "b0 = numpy.zeros(n1)\n", "\n", - "scale = numpy.sqrt(6/(nz1+1))\n", - "W1 = numpy.random.uniform(-scale, scale, nz1)\n", + "scale = numpy.sqrt(6/(n1+1))\n", + "W1 = numpy.random.uniform(-scale, scale, n1)\n", "b1 = 0." ] }, @@ -606,9 +640,19 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 13, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Initial precision: 59.2%\n", + "Initial recall: 88.5%\n", + "Initial F-score: 71.0%\n" + ] + } + ], "source": [ "# initial accuracy\n", "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", @@ -629,19 +673,30 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The optimization is the same as how we did it in lesson 5, except that now we have to update `W0`, `b0`, `W1`, and `b1` (in lesson 5, we only updated `W` and `b`):" + "The optimization is the same as how we did it in lesson 5, except that now we have to update `W0`, `b0`, `W1`, and `b1` (in lesson 5, we only updated `W` and `b`). Also, we give you a proper step size (`lr`) directly." ] }, { "cell_type": "code", - "execution_count": null, + "execution_count": 14, "metadata": { "scrolled": true }, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "10...20...30...40...50...60...70...80...90...100...110...120...130...140...150...160...170...180...190...200...210...220...230...240...250...260...270...280...290...300...310...320...330...340...350...360...370...380...390...400...410...420...430...440...450...CPU times: user 3min, sys: 9.24 s, total: 3min 9s\n", + "Wall time: 31.7 s\n" + ] + } + ], "source": [ - "# learning rate\n", - "lr = 5e-4\n", + "%%time\n", + "\n", + "# step size\n", + "lr = 1e-1\n", "\n", "# a variable for the change in validation loss\n", "change = numpy.inf\n", @@ -655,7 +710,7 @@ "# keep running if:\n", "# 1. we still see significant changes in validation loss\n", "# 2. iteration counter < 10000\n", - "while change >= 1e-5 and i < 10000:\n", + "while change >= 1e-6 and i < 10000:\n", " \n", " # calculate gradients and use gradient descents\n", " grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", @@ -680,13 +735,40 @@ " \n", " # print the progress every 10 steps\n", " if i % 10 == 0:\n", - " print(\"{}...\".format(i), end=\"\")\n", + " print(\"{}...\".format(i), end=\"\")" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "\n", + "Upon optimization stopped:\n", + " Iterations: 453\n", + " Validation loss w/o regularization: 0.2787043104606595\n", + " Validation loss w/ regularization: 0.024497566748684113\n", + " Validation precision: 0.8993288590604027\n", + " Validation recall: 0.8589743589743589\n", + " Validation F-score: 0.8786885245901639\n", + " Change in validation loss: 1.6184802410438597e-07\n" + ] + } + ], + "source": [ + "val_loss_reg = regularized_loss(images_train, labels_train, (W0, b0, W1, b1))\n", "\n", "print(\"\")\n", "print(\"\")\n", "print(\"Upon optimization stopped:\")\n", "print(\" Iterations:\", i)\n", - "print(\" Validation loss:\", val_loss)\n", + "print(\" Validation loss w/o regularization:\", val_loss)\n", + "print(\" Validation loss w/ regularization:\", val_loss_reg)\n", "print(\" Validation precision:\", score[0])\n", "print(\" Validation recall:\", score[1])\n", "print(\" Validation F-score:\", score[2])\n", @@ -702,9 +784,19 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 16, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Final precision: 91.8%\n", + "Final recall: 93.6%\n", + "Final F-score: 92.7%\n" + ] + } + ], "source": [ "# final accuracy\n", "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", @@ -718,7 +810,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Awesome! Compared to lesson 5, simply replacing the logistic regression model with a fully connected neural network model with $L=1$ improves the F-score from $86.4\\%$ to $91.6\\%$. The precision is now $92.8\\%$, meaning whenever our model predicts a casting part being defective, $92.8\\%$ chance it's correct. And the recall is now $90.4\\%$, which says that our model misses around $10\\%$ of defective parts." + "Awesome! Compared to lesson 5, simply replacing the logistic regression model with a fully connected neural network model with $L=1$ improves the F-score from $86.6\\%$ to $92.7\\%$. The precision is now $91.8\\%$, meaning whenever our model predicts a casting part being defective, $91.8\\%$ chance it's correct. And the recall is now $93.4\\%$, which says that our model misses around $6.6\\%$ of defective parts." ] }, { @@ -756,9 +848,171 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 17, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "text/html": [ + "\n", + "\n", + "\n", + "\n", + "\n" + ], + "text/plain": [ + "" + ] + }, + "execution_count": 17, + "metadata": {}, + "output_type": "execute_result" + } + ], "source": [ "# Execute this cell to load the notebook's style sheet, then ignore it\n", "from IPython.core.display import HTML\n", @@ -783,7 +1037,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.9.5" + "version": "3.9.6" } }, "nbformat": 4, From 3b77efa823d1564eb5362db26c536525fe2ab287 Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Fri, 6 Aug 2021 02:55:59 -0400 Subject: [PATCH 4/8] add learning rate explanation and clear output of lesson 7 --- .../7_The_First_Deep_Neural_Network.ipynb | 263 ++---------------- notebooks_en/Appendix.ipynb | 12 +- 2 files changed, 34 insertions(+), 241 deletions(-) diff --git a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb index 835bdc0..96fe342 100644 --- a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb +++ b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb @@ -30,7 +30,7 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -48,7 +48,7 @@ }, { "cell_type": "code", - "execution_count": 2, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -64,7 +64,7 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -89,7 +89,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -369,7 +369,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -406,7 +406,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -446,7 +446,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -487,7 +487,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -504,7 +504,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -529,7 +529,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -549,18 +549,9 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Gradients of W0 are zeros: True\n", - "Gradients of W1 are zeros: False\n" - ] - } - ], + "outputs": [], "source": [ "grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", "print(\"Gradients of W0 are zeros:\", numpy.allclose(grads[0], 0.))\n", @@ -610,7 +601,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": null, "metadata": {}, "outputs": [], "source": [ @@ -640,19 +631,9 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Initial precision: 59.2%\n", - "Initial recall: 88.5%\n", - "Initial F-score: 71.0%\n" - ] - } - ], + "outputs": [], "source": [ "# initial accuracy\n", "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", @@ -678,20 +659,11 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": null, "metadata": { "scrolled": true }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "10...20...30...40...50...60...70...80...90...100...110...120...130...140...150...160...170...180...190...200...210...220...230...240...250...260...270...280...290...300...310...320...330...340...350...360...370...380...390...400...410...420...430...440...450...CPU times: user 3min, sys: 9.24 s, total: 3min 9s\n", - "Wall time: 31.7 s\n" - ] - } - ], + "outputs": [], "source": [ "%%time\n", "\n", @@ -740,26 +712,9 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "\n", - "Upon optimization stopped:\n", - " Iterations: 453\n", - " Validation loss w/o regularization: 0.2787043104606595\n", - " Validation loss w/ regularization: 0.024497566748684113\n", - " Validation precision: 0.8993288590604027\n", - " Validation recall: 0.8589743589743589\n", - " Validation F-score: 0.8786885245901639\n", - " Change in validation loss: 1.6184802410438597e-07\n" - ] - } - ], + "outputs": [], "source": [ "val_loss_reg = regularized_loss(images_train, labels_train, (W0, b0, W1, b1))\n", "\n", @@ -784,19 +739,9 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Final precision: 91.8%\n", - "Final recall: 93.6%\n", - "Final F-score: 92.7%\n" - ] - } - ], + "outputs": [], "source": [ "# final accuracy\n", "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", @@ -848,171 +793,9 @@ }, { "cell_type": "code", - "execution_count": 17, + "execution_count": null, "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "\n", - "\n", - "\n", - "\n", - "\n" - ], - "text/plain": [ - "" - ] - }, - "execution_count": 17, - "metadata": {}, - "output_type": "execute_result" - } - ], + "outputs": [], "source": [ "# Execute this cell to load the notebook's style sheet, then ignore it\n", "from IPython.core.display import HTML\n", diff --git a/notebooks_en/Appendix.ipynb b/notebooks_en/Appendix.ipynb index 82b67d8..4687010 100644 --- a/notebooks_en/Appendix.ipynb +++ b/notebooks_en/Appendix.ipynb @@ -107,6 +107,16 @@ "In lesson 5, we used the example of take-home exercises, quizzes, and final exams to explain the concepts of training, validation, and test datasets. We can see the optimization of a model is indeed similar to training a student to some degree." ] }, + { + "cell_type": "markdown", + "id": "66f3e93d", + "metadata": {}, + "source": [ + "##### Learning rate\n", + "\n", + "The *step size* in gradient-descent-based optimization methods is called learning rate." + ] + }, { "cell_type": "markdown", "id": "dbdcb6fa", @@ -147,7 +157,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.9.5" + "version": "3.9.6" } }, "nbformat": 4, From d1c649ef09d66534d90508f08cda24460a3077cc Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Mon, 9 Aug 2021 13:27:34 -0400 Subject: [PATCH 5/8] lesso5: record loss hist, hist plot, run a fixed num. iter. --- .../5_Multiple_Logistic_Regression.ipynb | 160 ++++++++++-------- 1 file changed, 94 insertions(+), 66 deletions(-) diff --git a/notebooks_en/5_Multiple_Logistic_Regression.ipynb b/notebooks_en/5_Multiple_Logistic_Regression.ipynb index 148ffbc..20d2989 100644 --- a/notebooks_en/5_Multiple_Logistic_Regression.ipynb +++ b/notebooks_en/5_Multiple_Logistic_Regression.ipynb @@ -875,13 +875,15 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## Training/optimization\n", + "## Optimizing the model up to when validation loss stops improving\n", "\n", - "In the optimization loop (using gradient descent), we'll monitor the progress of training using the validation dataset. If the validation loss is not changing much, we'll stop the training. The loss we want to monitor is the original loss. On the other hand, when calculating the gradients, we use the regularized loss against the training data.\n", + "In lesson 4, we introduced the concept of overfitting -- overfitting happens when a model predicts accurately against training data but is not so good at data not in the training dataset, such as validation and test data. In other words, after running the gradient descent up to some iteration, if we keep running it, though training loss will keep going down, the validation loss may not. The validation loss may reach a plateau or even start to degrade. \n", "\n", - "Here, we only use validation loss to control the optimization. In real applications, it's common to use a combination of validation loss, accuracy, and other metrics to control the optimization.\n", + "Monitoring the change of validation loss in real-time is an excellent way to stop the optimization early before overfitting happens. However, to avoid the hassle of programming, we adopt a more naive strategy: we run the optimization with a fixed number of iterations, record the history of validation loss, and use the parameters where the lowest validation loss happens as the final solution.\n", "\n", - "Also, the step size (the variable `lr`) is `1e-2`. The step size affects whether we can and how fast we'll find the solution. Finding a proper step size requires hyperparameter tuning mentioned previously. Here we just give you a proper value we found as hyperparameter tuning is out of the scope of this lesson." + "The step size (the variable `lr`) is `1e-3`. It affects whether we can and how fast we'll find the solution. Finding a proper step size requires hyperparameter tuning. We give you a value directly because hyperparameter tuning is out of the scope of this lesson.\n", + "\n", + "(Note, the following code cell will take about 2 minutes to finish with a personal desktop.)" ] }, { @@ -895,8 +897,9 @@ "name": "stdout", "output_type": "stream", "text": [ - "10...20...30...40...50...60...70...80...90...100...110...120...130...140...150...160...170...180...190...200...210...220...230...240...250...260...270...280...290...300...310...320...330...340...350...360...370...380...390...400...410...420...430...440...450...460...470...480...490...500...510...520...530...540...550...560...570...580...590...600...610...620...630...640...650...660...CPU times: user 34.2 s, sys: 412 ms, total: 34.6 s\n", - "Wall time: 5.77 s\n" + "0...100...200...300...400...500...600...700...800...900...1000...1100...1200...1300...1400...1500...1600...1700...1800...1900...2000...2100...2200...2300...2400...2500...2600...2700...2800...2900...3000...3100...3200...3300...3400...3500...3600...3700...3800...3900...4000...4100...4200...4300...4400...4500...4600...4700...4800...4900...5000...5100...5200...5300...5400...5500...5600...5700...5800...5900...6000...6100...6200...6300...6400...6500...6600...6700...6800...6900...7000...7100...7200...7300...7400...7500...7600...7700...7800...7900...8000...8100...8200...8300...8400...8500...8600...8700...8800...8900...9000...9100...9200...9300...9400...9500...9600...9700...9800...9900...\n", + "CPU times: user 10min 9s, sys: 6.73 s, total: 10min 16s\n", + "Wall time: 1min 42s\n" ] } ], @@ -904,44 +907,43 @@ "%%time\n", "\n", "# step size\n", - "lr = 1e-2\n", - "\n", - "# a variable for the change in validation loss\n", - "change = numpy.inf\n", + "lr = 1e-3\n", "\n", - "# a counter for optimization iterations\n", - "i = 0\n", + "# variables to store parameters at when the validation loss is at its lowest\n", + "best_val_loss = numpy.inf\n", + "best_iter = 0\n", + "best_w = numpy.zeros_like(w)\n", + "best_b = 0.\n", "\n", - "# a variable to store the validation loss from the previous iteration\n", - "old_val_loss = 1e-15\n", + "# variables to store the history of training and validation loss\n", + "hist_train_loss = []\n", + "hist_val_loss = []\n", "\n", - "# keep running if:\n", - "# 1. we still see significant changes in validation loss\n", - "# 2. iteration counter < 10000\n", - "while change >= 1e-6 and i < 10000:\n", + "# run for 10000 iterations\n", + "for i in range(10000):\n", " \n", " # calculate gradients and use gradient descents\n", " grads = gradients(images_train, labels_train, (w, b))\n", - " w -= (grads[0] * lr)\n", - " b -= (grads[1] * lr)\n", + " w = w - grads[0] * lr\n", + " b = b - grads[1] * lr\n", " \n", - " # validation loss\n", - " val_loss = model_loss(images_val, labels_val, (w, b))\n", + " # save losses for future investigation\n", + " hist_train_loss.append(model_loss(images_train, labels_train, (w, b)))\n", + " hist_val_loss.append(model_loss(images_val, labels_val, (w, b)))\n", " \n", - " # calculate f-scores against the validation dataset\n", - " pred_labels_val = classify(images_val, (w, b))\n", - " score = performance(pred_labels_val, labels_val)\n", - "\n", - " # calculate the chage in validation loss\n", - " change = numpy.abs((val_loss-old_val_loss)/old_val_loss)\n", - "\n", - " # update the counter and old_val_loss\n", - " i += 1\n", - " old_val_loss = val_loss\n", + " # if the validation loss is lower than the current best, save it\n", + " if hist_val_loss[-1] < best_val_loss:\n", + " best_val_loss = hist_val_loss[-1]\n", + " best_iter = i\n", + " best_w = w\n", + " best_b = b\n", " \n", - " # print the progress every 10 steps\n", - " if i % 10 == 0:\n", - " print(\"{}...\".format(i), end=\"\")" + " # print something every 100 iterations so we know it's not dead\n", + " if i % 100 == 0:\n", + " print(\"{}...\".format(i), end=\"\")\n", + "\n", + "# just to add a new line\n", + "print()" ] }, { @@ -953,39 +955,65 @@ "name": "stdout", "output_type": "stream", "text": [ - "\n", - "\n", - "Upon optimization stopped:\n", - " Iterations: 666\n", - " Validation loss w/o regularization: 0.5188273788049017\n", - " Validation loss w/ regularization: 0.04127987992406643\n", - " Validation precision: 0.8951048951048951\n", - " Validation recall: 0.8205128205128205\n", - " Validation F-score: 0.8561872909698997\n", - " Change in validation loss: 9.484651848988673e-07\n" + "The best validation loss: 0.4230061110360034\n", + "It happened at iteration 1472\n" ] } ], "source": [ - "val_loss_reg = regularized_loss(images_train, labels_train, (w, b))\n", - "\n", - "print(\"\")\n", - "print(\"\")\n", - "print(\"Upon optimization stopped:\")\n", - "print(\" Iterations:\", i)\n", - "print(\" Validation loss w/o regularization:\", val_loss)\n", - "print(\" Validation loss w/ regularization:\", val_loss_reg)\n", - "print(\" Validation precision:\", score[0])\n", - "print(\" Validation recall:\", score[1])\n", - "print(\" Validation F-score:\", score[2])\n", - "print(\" Change in validation loss:\", change)" + "print(\"The best validation loss:\", best_val_loss)\n", + "print(\"It happened at iteration {}\".format(best_iter))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "We didn't use the validation F-score to stop the optimization because improving the predicted probability does not necessarily improve the predicted labels. For example, suppose the probabilities predicted by the logistic model during the last three iterations are $0.1$, $0.2$, and $0.3$. In that case, the predicted labels are still $0$, $0$, and $0$ for the three iterations. Hence monitoring F-score (which relies on predicted labels) will cause the optimization to stop too early while the model is still improving." + "A good thing of recording the history of loss is that we can visualize and inspect them:" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAA4ZUlEQVR4nO3dd3xV9f348dc7dyS5uRlkQCABEgREUFlhKA60Wmcd1FbRKpS62trW2vHV1lqt+v122P5s62ipexXrog7U1oE4WiUMBxsDQhghg+x5k8/vj3MSbsLNIOTmJve8n4/Hfdyz7/tzb/J5n/M553yOGGNQSinlXDGRDkAppVRkaSJQSimH00SglFIOp4lAKaUcThOBUko5nCYCpZRyOE0Eqh0ReVVEFvT1sv1FRG4VkSfs4VEiUi0iru6W7eVnrRORub1dv4vtLheRK/t6u32pqxi7+97VwOOOdADq8IlIddCoD2gAmu3xa4wxT/Z0W8aYs8KxbE+IyKXAucaYS/tie8aYHYC/L7YlIo8AhcaYm4O2P6kvth1tevq9i8hC4EpjzAlhD0p1SRNBFDDGtP3Tich2rH+uNzouJyJuY0ygP2M7RGcDyyIdhBocBsHf86ChTUNRTETmikihiPyPiOwFHhaRISLysogUi8h+ezg7aJ22Q34RWSgi74nIXfay20TkrF4umysiK0SkSkTeEJF7g5tlRCQGOB14TUReE5HrOpTlYxGZZw//UUR2ikiliKwSkRM7KX+OiBgRcQfF8I4dw7+B9A7LPyMie0Wkwo51kj39auAy4Kd2k8dL9vTtInKaPRwrIneLyG77dbeIxHb4HX4kIvtEZI+IfLOHv2GMiNwsIl/Y6z4mIsn2vDgReUJESkWkXERWisiwoN+jwC7rNhG5LGibi0Rkg/07vS4io+3pIiL/z/6cChH5RESO7iK80SLyvv0Z/xKR9E6+94NiEZGjgL8Ax9nfabm9bLJdxmK7zDfbfxut23nfjrEMuF1EykTkmKCyDRWROhHJ6Mn3qyyaCKJfJpAKjAauxvrNH7bHRwF1wD1drD8L2IRVaf4WeFBEpBfLPgV8BKQBtwKXd1h3JlBgjCmxl53fOkNEJtrxvmJPWglMscv1FPCMiMR1UYZWTwGr7PhuBzqe33gVGAcMBVYDTwIYYxbbw781xviNMV8Jse2fA7PtuCbb5bk5aH4mkAxkAd8C7hWRIT2IeaH9OgUYg9Xk0vp7LbC3ORLre70WqBORBOBPwFnGmETgeGAtgIhcAPwMmAdkAO8Cf7e392XgJGA8kAJcDJR2EdulwDexvi8v8OOOC3QWizFmgx3vf+zvNMVe5c92mcYAJwNX2J/RahZQYH/mr4AlwDeC5s8H3jDGFHcRt+rIGKOvKHoB24HT7OG5QCMQ18XyU4D9QePLsZqWwKqAtgbN8wEGyDyUZbESTgDwBc1/AngiaPx24Bf2cCJQA4y2x+8EHuqiDPuByfbwra3bBXLsGNxBMSQErfdUcAwdtplir5tsjz8C3NHFd/05cHbQvDOA7UG/Qx3gDpq/D5jdyWcHf69vAt8Jmnck0GSXaRHwAXBsh/UTgHLgq0B8h3mvAt8KGo8BarES7anAZqyEFtPN39ly4Oag8e8Ar4X43ruKZSHwXtC4C+v81sSgadcAy4OW39FhG7OAna3xAvnA1yP9fzjYXnpEEP2KjTH1rSMi4hORv9qH3ZXACiBFOr/CY2/rgDGm1h7s7ERgZ8uOAMqCpoH1zxus7fyAMaYKa+//EnveJdh753YZfmQ3bVTYTQrJdGjmCWEEVsKrCZr2RdA2XSLyaxH53P5ettuzuttu8Pa/CBr/wp7WqtS0b8+upWcnskNt1w0MAx4HXgeW2M1RvxURj13Gi7H2uPeIyCsiMsFefzTwR7spqRwoAwTIMsa8hXW0cS9QJCKLRSSpi9j2Bg2HLE83sXSUjnVk0bG8WUHj7f5ujDEfYu00nGxvdyzwYhcxqxA0EUS/jt3L/ghrr3KWMSYJqykArMogXPYAqSLiC5o2snVARDKB4VjNMa3+DswXkeOAeOBte9kTgf8Bvg4MMVaTQkUP4t8DDLGbKlqNChq+FDgfOA0rseS0hme/d9dN726sSjZ427u7WacnQm03ABQZY5qMMbcZYyZiNbmci9WUgjHmdWPM6Vjf60bgb/b6O7GuJEsJesUbYz6w1/uTMWY6MAmriegnh1uALmLp+J2WYB3tdCzvruDNhfiIR7Gahy4Hng3e8VE9o4nAeRKxminKRSQV+GW4P9AY8wXWIfutIuK1K/fgdvazsZoVgv/Jl2FVCL8CnjbGtATFHwCKAbeI3AJ0tdfaMYbb7BhO6BBDIlazRClWs9b/dthEEVa7dWf+DtwsIhn2SdNbsJq/DtffgR+KdaLbb8f1tDEmICKniMgx9tFcJVYl2iwiw0TkPDvpNQDVHLic+C/ATXLgRHiyiHzNHp4hIrNExIO1l10ftF6vdBNLEZAtIl4AY0wz8A/gThFJtE9i30D33+PjwIVYyeCxw4nXqTQROM/dWHvYJcB/gdf66XMvA47DqmjvAJ7GqhggxGWjxpgG4HmsPfSngma9jtXOvRmr2aCeg5uZOnMpVptyGVYCDK40HrO3twtYj/XdBHsQmGg3qSwNse07sBLNJ8CnWEc3d/Qwrq48hFXRrQC2YZX3e/a8TOBZrCSwAXgHq9KMwTry241V1pOx2vAxxrwA/AarOakS+AxovborCWtvfT/Wd1EK3HWY8XcaC/AWsA7YKyIl9rTvYSWhAuA9rN/+oa4+wBhTiPV9G6yT3+oQSfudMKX6h4g8jdVMcDtWW/MRxpiKyEalBisReQjYbYJu+FM9p4lA9QsRmYG1R7gN6zLFpVhHCLuArxpj7o9cdGowE5EcrMtjpxpjtkU2msFJ7yxW/SUTq6knDSgEvm2MWWPP0ySgekVEbgd+CPyfJoHe0yMCpZRyOD1ZrJRSDjfomobS09NNTk5OpMNQSqlBZdWqVSXGmJB9MIU1EYjImcAfsW4df8AY8+sQy8zFuqTRA5QYY07uaps5OTnk5+f3eaxKKRXNROSLzuaFLRHYN7nci9WjZCGwUkReNMasD1omBbgPONMYs0NEhoYrHqWUUqGF8xzBTKxOyAqMMY1YvQSe32GZS4HnjfUgC4wx+8IYj1JKqRDCmQiyaH/HZyHtO48Cqy+TIWL1a79KRK4ItSERuVpE8kUkv7hYe5dVSqm+FM5EEKoTsI7XqrqB6cA5WN32/kJExh+0kjGLjTF5xpi8jAx93oRSSvWlcJ4sLiSoh0kgm4N7YyzEOkFcA9SIyAqsh3psDmNcSimlgoTziGAlMM7uNdGL1ad8x37C/wmcKCJuu4viWVidZymllOonYTsisLvJvQ6rt0gX1hOm1onItfb8vxhjNojIa1g9NrZgXWL6WbhiUkopdbBB18VEXl6e6dV9BEXrYd3zMPMa8Ot5BqWUs4jIKmNMXqh5zuliomQTrPgd1OhVR0opFcw5iaD1kbxtD7pSSikFjkoEdlHNYT15Tymloo5zEkGMfUTQoolAKaWCOScRaNOQUkqF5JhEUFoXAKCmoTHCkSil1MDimESwZV8NAGVVdRGORCmlBhbHJAKJse6da2kJRDgSpZQaWByTCFpPFptmPUeglFLBHJMIJMYqqtEjAqWUasc5iaDtqqHB1aWGUkqFm3MSgctKBHqOQCml2nNOIhC9oUwppUJxTCJoO1msiUAppdpxTCKIidE7i5VSKhTHJILWq4ZamvWIQCmlgjkmEWDfUKaXjyqlVHuOSQStTUOiTUNKKdWOYxKB6MlipZQKyTGJQK8aUkqp0ByTCGLcXmuguSmygSil1ADjmESAJx6AmOb6CAeilFIDi+MSgTTVRjgQpZQaWJyTCFxemo0QE9AH0yilVDDHJAKXK4Za4jQRKKVUB45JBDEi1OPVcwRKKdWBgxIB1BkvroCeI1BKqWBhTQQicqaIbBKRrSJyY4j5c0WkQkTW2q9bwhgLVfhwNVWH6yOUUmpQcodrw2I9AOBe4HSgEFgpIi8aY9Z3WPRdY8y54YqjlUuEMpPI8IaycH+UUkoNKuE8IpgJbDXGFBhjGoElwPlh/LwuxYhQRhJeTQRKKdVOOBNBFrAzaLzQntbRcSLysYi8KiKTQm1IRK4WkXwRyS8uLu5VMCJQZhLxNpb3an2llIpW4UwEEmJaxyfHrwZGG2MmA38GlobakDFmsTEmzxiTl5GR0atg3C6racgbqIJAY6+2oZRS0SiciaAQGBk0ng3sDl7AGFNpjKm2h5cBHhFJD0cwXlcMpSRbI7Ul4fgIpZQalMKZCFYC40QkV0S8wCXAi8ELiEimiIg9PNOOpzQcwXjcMewxqdZIxa5wfIRSSg1KYbtqyBgTEJHrgNcBF/CQMWadiFxrz/8LcBHwbREJAHXAJcaYjs1HfcLriqHQ2M1K5V/AyBnh+BillAoPY6AlAC5Pn286bIkA2pp7lnWY9peg4XuAe8IZQyuvK4Zdxm51Kt/RHx+plFKdMwYaKqGmxHrV2u81xVBbGjRcAjWl1vucH8CpN/d5KGFNBANJTIzQGBNPrTsFX8XO7ldQSqlDYQw0VluVd3VxUCVefKAi71jpt3TyfBSvH3xpkJAOiSMgczIkpMHo48MSumMSAYDHFUO5dzi+sm2RDkUpNRi0NENtGdTsa1/B1xTb00qget+BvffOOrX0+q1K3ZcOyVkwYrI13DotIcOq6Fun2d3m9xdHJQKvO4aiuBxG7Fsd6VCUUpHSVBdUee87ULFXh6jga0s5+Kp3QFxW5e3PsN7Tx1sVeEIG+IfaFbs97ksHT1y/F/NQOCoReFwx7Iody9SyV60f3d+7exKUUgNISwvUl9uVeVDF3lkF39hJf2Nev12BZ0DqGBg50x4falXqbRV8BsSlQEz09NnpqEQQ645hp3eMNVL0KfhPjWxASqnONVRDdZFVubd77zCtZp91NU1HEmO3s9uVd1aeXZmnB1Xw9l69Lx28vv4v4wDhqETgcQnbPWMBgZ0fwRGaCJTqV4FGq+LutIIvPjCtqebg9SXGqsD9Q8E/DIZNsiv3oUGVvF3B+1IhxtX/ZRyEHJUIvO4YKkwCjJgCBcth7kE9YyulDlVLC9SVhd5b7zitbn/obcQPsSr2hAzImm4Nt1b2be/DtHIPE0clAo8rhqbmFhgzFz74M9RXQFxypMNSamAKNEDVXqsCr9pjDVftPbiCrykG03zw+h7fgUo8fRzknBC6gk/IAHds/5dPtXFUIojzuKhraoYjz4b3/h+sfxGmXR7psJTqX4HGAxV5WwUfVNG3jteF6LI9xn2gGSZxOAyffGBvvWMFH+vv/7KpXnFUIkiIdVNR1wTZsyBtLKx9UhOBih7NTfbe+t4OlfseqCo6MByq00VxWZV3YiYMyYFRs62KPjEz6DUc4lOj6moZZXFWIvC62FNeZz2cYPpC+NfNsONDGDUr0qEp1bmWZqv55aC999YK3h6vKeaga95bT64mZkJyNmTn2RX8sKCKfrh1dY22vTuWoxKBz+umttFuy8xbBO/dDW/+Cha+bCUHpfpTS4u1d95V80zVXusqG9PSYWWx2tZbK/IRU9vvube+J2RoBa+65ahEkBDrorrBvt7Ym2B13vTy9bDqYSsxKNUXWq+i6a6Cry4KfZLVl36gMs88un3FHlzBh6EXSuVMDksEbmobg248mb4Q1i+F126CYcdo19Sqa8ZYlz9W7em8eaZqr9VGH+oGp/jUA80yQ486uHJvPenq9vZ/2ZSjOSsReF00NRsaAy143TFWc9BXH4QHvgRPfR2+8RxkTYt0mKq/GWN1UdBV+3trBd8c4jGncSkHKvj0E9tX8H67ucY/bMD3N6Ocy1GJwOe1ilvTEMDbuteVkA7feB4evwAe/Qqcfw9MujByQaq+066CD6rMQzXTNDccvH5ssn1SNRNGH9d+zz34ipp+7ilSqb7mqETgj7UTQWOAIQlBh99pR8Cif8HTl8EzC2HTa3D6bdY/uRp4Wpto2q6Dt99DjQfqD17f6z9QqY+c2X7PPbiC9yb0f9mUigBHJQJfrHX1RNuVQ8GShsOi1+Gd31hXE214CWZfCzOusuap8Gtrg9/becXeOh5qD96beKASHznz4D13f6a1hx+b2P9lU2oAc1QiaD0iqKrv5KlALo91JdHk+fDmbfDuH+D9P8GEc+DoeTD2dEf3UNgrrZV7WxfB++xugfcd6J6gtbOxmuLum2hGzm5/k5M/U/fglTpMjkoEKT6rOai8tpNE0CrtCPj6Y1BWAB8uhk+fsa4u8visOy5zToTRc6xL+5xW+RgDDVXWAzvqyqynN9WW2c9YtSv3dhV+cejH8bV7sMdQyJhgXxc//OCKXpOvUmHlrEQQb1133W0iaJU6Bs76NXz5DvjiPdj4Cmx71zpaAECsZTKPhtQjYMhoSBkFKaMH9h6qMVbbeX2l9fDs+kpoqAgar7Aq9zq7gq/d377i7+w5q2390NiV+7CjDzyxqbVzsdYug+OHaFcFSg0QzkoEPjsR1PUwEbRyua0eS8fMtcari6HwI9j7mfWAm72fWkmi47Xj7nj7wRhp1ntskpUcvAnW0UXruzvWuvtTXNZ7jNsejrHeTbPVzUBL4MB727RmqzmlqR6aaq0KvqnWHq9rPy244u+sMm8lLqvLX1+adf172hHgm3Fg3JfWfr4v1arc9Q5tpQYdRyWCpDgPIlBeG+Ja8EPhz7DOG0w458C05oB1IrP8CyjfYbV715TYzSYl1nD5TqtCbqyx3kNdk3443HHWpYzueOvd47OuXffEQ2ym9VzV2CTrZGlckjUcl2y/J3V4T9ZKXSmHcFQiiIkRkuM9PW8aOhQuN6SMtF491dxkJYXmRmvPvnUvP3hv37R0OFoIPmJwW0cNLq9V+WtTi1KqFxyVCMA6T3DITUPh4vJAfEqko1BKOZzjdiFTfN7DbxpSSqko4sBEEKamIaWUGqQclwhSfV7KavSIQCmlWoU1EYjImSKySUS2isiNXSw3Q0SaReSicMYDMDQpjn1V9Rhjul9YKaUcIGyJQERcwL3AWcBEYL6ITOxkud8Ar4crlmBDE2Npajbs1+YhpZQCwntEMBPYaowpMMY0AkuA80Ms9z3gOWBfGGNpMyzJ6hN+X1WIXimVUsqBwpkIsoCdQeOF9rQ2IpIFXAj8pasNicjVIpIvIvnFxcWHFdTQpFgAiipDdG6mlFIOFM5EEOq21I4N83cD/2NMqAe3Bq1kzGJjTJ4xJi8jI+OwghqWaB8RVOoRgVJKQXhvKCsEgm+zzQZ2d1gmD1giVlcG6cDZIhIwxiwNV1CtRwT7qvSIQCmlILyJYCUwTkRygV3AJcClwQsYY3Jbh0XkEeDlcCYBgDiPi6Q4N0V6RKCUUkAYE4ExJiAi12FdDeQCHjLGrBORa+35XZ4XCKesIT527a+L1McrpdSAEta+howxy4BlHaaFTADGmIXhjCXYyCHxbC+t6a+PU0qpAc1xdxYDjEr1saOsVm8qU0opnJoI0nzUN7VQXK0njJVSypGJYGSq9QzcnWW1EY5EKaUiz5GJYFRbItATxkop5chEkJUSjwh6wlgppXBoIojzuBiV6mPLvupIh6KUUhHnyEQAMG5oIluKqiIdhlJKRZxzE8EwP9tKamhqbol0KEopFVGOTQTjh/lpajZsL9HzBEopZ3NsIhg3NBGAzUV6nkAp5WyOTQRjh/qJEdik5wmUUg7n2EQQ53Exdqifz3ZVRDoUpZSKKMcmAoBjslL4pLBc+xxSSjmaoxPB5JHJlFQ3srtCn02glHIuRyeCY7NTAPhkZ3lE41BKqUhydCKYkJmIO0b4RM8TKKUczNGJIM7j4qjhSazZsT/SoSilVMQ4OhEAzMhJZc2OchoCzZEORSmlIqJHiUBEEkQkxh4eLyLniYgnvKH1j9ljUmkItPDxTm0eUko5U0+PCFYAcSKSBbwJfBN4JFxB9aeZuamIwIcFpZEORSmlIqKniUCMMbXAPODPxpgLgYnhC6v/pPi8HDkskQ+3lUU6FKWUiogeJwIROQ64DHjFnuYOT0j9b/aYNPK/KNPzBEopR+ppIrgeuAl4wRizTkTGAG+HLap+dsLYdOqbWsjfrlcPKaWcp0d79caYd4B3AOyTxiXGmO+HM7D+dPzYNLzuGN7euI85Y9MjHY5SSvWrnl419JSIJIlIArAe2CQiPwlvaP3H53Uze0wab23aF+lQlFKq3/W0aWiiMaYSuABYBowCLg9XUJFwypEZFBTX8IU+0F4p5TA9TQQe+76BC4B/GmOagKjqsvOUI4cC8PZGPSpQSjlLTxPBX4HtQAKwQkRGA5XdrSQiZ4rIJhHZKiI3hph/voh8IiJrRSRfRE44lOD7Uk56AmOH+nl9XVGkQlBKqYjoUSIwxvzJGJNljDnbWL4ATulqHRFxAfcCZ2HdczBfRDree/AmMNkYMwVYBDxwqAXoS+ccM5wPt5Wyr0q7pVZKOUdPTxYni8gf7L32fBH5PdbRQVdmAluNMQXGmEZgCXB+8ALGmGpz4KkwCUS4uemcY4fTYuC1z/ZGMgyllOpXPW0aegioAr5uvyqBh7tZJwvYGTReaE9rR0QuFJGNWDeqLQq1IRG5ujUJFRcX9zDkQzd+WCLjh/l5+ZM9YfsMpZQaaHqaCI4wxvzS3rsvMMbcBozpZh0JMe2gPX5jzAvGmAlYJ6JvD7UhY8xiY0yeMSYvIyOjhyH3zrnHjmDl9jL26lPLlFIO0dNEUBd8IldE5gB13axTCIwMGs8Gdne2sDFmBXCEiET0jq6vTB6BMfDCml2RDEMppfpNTxPBtcC9IrJdRLYD9wDXdLPOSmCciOSKiBe4BHgxeAERGSsiYg9PA7xARLsBzU1PYGZOKs/k79SH2iulHKGnVw19bIyZDBwLHGuMmQqc2s06AeA64HVgA/APu5+ia0XkWnuxrwKficharCuMLjYDoPb9Wl42BSU1rNS+h5RSDiC9rXdFZIcxZlQfx9OtvLw8k5+fH9bPqG0MMPPONzljUia///rksH6WUkr1BxFZZYzJCzXvcB5VGepkcFTwed18ZfJwln26h4q6pkiHo5RSYXU4iSDiTTjhdNms0dQ1NfNM/s7uF1ZKqUGsy0QgIlUiUhniVQWM6KcYI+LorGRm5qby8PvbCTS3RDocpZQKmy4TgTEm0RiTFOKVaIyJmieUdWbRnFx2ldfx7/Xa/5BSKnodTtNQ1Dt94jBGpsbz4HvbIh2KUkqFjSaCLrhihG8en0v+F/vJ364Pt1dKRSdNBN2YP3MU6X4vf3xzS6RDUUqpsNBE0I14r4urTxrDu1tKWPWF3mCmlIo+mgh64BuzR5OaoEcFSqnopImgB3xeN1edOIYVm4tZqecKlFJRRhNBDy04fjTDkmK585UN2hmdUiqqaCLoIZ/XzY++fCRrd5bzyqf64BqlVPTQRHAIvjotmwmZifz2tU00BJojHY5SSvUJTQSHwBUj/Pyco9hRVssj72+PdDhKKdUnNBEcohPHZXDaUcO4+40t7Crv7iFtSik18Gki6IVbz5sIwC//uS7CkSil1OHTRNAL2UN8XH/aON7YUMS/1u2NdDhKKXVYNBH00qITcpmQmcgt/1ynD69RSg1qmgh6yeOK4TdfPZbi6gZue1GbiJRSg5cmgsMweWQK3z1lLM+v2cWrem+BUmqQ0kRwmL536liOzkriZy98yr6q+kiHo5RSh0wTwWHyuGL4f1+fQk1jMz/6x8c0t2j3E0qpwUUTQR8YNyyR286bxLtbSrjnra2RDkcppQ6JJoI+csmMkcybmsXdb27mvS0lkQ5HKaV6TBNBHxER7rjwaMYN9fODJWvYrXcdK6UGCU0EfcjndXPfZdNpDLRw5aP51DQEIh2SUkp1SxNBHxs71M+fL53Kxr2VXP/0Wlr05LFSaoALayIQkTNFZJOIbBWRG0PMv0xEPrFfH4jI5HDG01/mHjmUW86dyL/XF/Gb1zdGOhyllOqSO1wbFhEXcC9wOlAIrBSRF40x64MW2wacbIzZLyJnAYuBWeGKqT8tOD6Hz4tr+Os7BQxPimPhnNxIh6SUUiGFLREAM4GtxpgCABFZApwPtCUCY8wHQcv/F8gOYzz9SkT45VcmsreynltfWk+yz8OFU6OmeEqpKBLOpqEsYGfQeKE9rTPfAl4NNUNErhaRfBHJLy4u7sMQw8vtiuHP86dy3Jg0fvzMJ7yxvijSISml1EHCmQgkxLSQZ05F5BSsRPA/oeYbYxYbY/KMMXkZGRl9GGL4xXlc/G1BHkePSOI7T61mxebBk8iUUs4QzkRQCIwMGs8GdndcSESOBR4AzjfGlIYxnojxx7p5+JszOSLDz5WP5fP2xn2RDkkppdqEMxGsBMaJSK6IeIFLgBeDFxCRUcDzwOXGmM1hjCXiUhO8/P2qWRw5LJGrH8/XB9oopQaMsCUCY0wAuA54HdgA/MMYs05ErhWRa+3FbgHSgPtEZK2I5IcrnoEgxefliStnMWlEMt95cjUvfnzQAZJSSvU7MWZw3fCUl5dn8vMHd76oqm/iW4/k89H2Mm4+5yiuPHFMpENSSkU5EVlljMkLNU/vLI6AxDgPj31rJmcfk8kdr2zgVy+t1zuQlVIRo4kgQuI8Lu6ZP41Fc3J56P1tXPf31dQ2at9ESqn+F84bylQ3YmKEW74ykREpcdy5bAPbSmpZfPl0Rqb6Ih2aUspB9IhgALjyxDE8vHAGu/bXct497/HBVn2egVKq/2giGCDmHjmUf153Aun+WC5/6CMWr/hczxsopfqFJoIBJDc9gRe+O4fTjxrG/y7byKJHV1JS3RDpsJRSUU4TwQDjj3Vz/zemcfv5k/jg81LO+uO7vK9NRUqpMNJEMACJCJcfl8M/vzuH5HgP33jwQ/5v2Qbqm5ojHZpSKgppIhjAjhqexIvXzeGSGaP464oCzvnTu6zesT/SYSmloowmggHO53Xzf/OO4fFvzaS+qYWL7v+A/9WjA6VUH9JEMEicOC6D164/kUtmjmLxigLOuHsFyzdpL6ZKqcOniWAQSYzz8L8XHsNTV87CFSMsfHgl1z6+il3ldZEOTSk1iGkiGISOH5vOaz84iZ+eeSTvbC7mtN+/w33Lt2pzkVKqVzQRDFJedwzfmTuWN350MieNT+e3r23iS79/hxfWFOqNaEqpQ6KJYJDLSonnr5fn8dSVsxiS4OGHT3/MV+55T+89UEr1mCaCKHH82HRe/O4J/PGSKZTXNnHZAx9y+YMfsuoLvdxUKdU1fTBNFGoINPP4f77gvuWfU1bTyInj0vnBl8aRl5Ma6dCUUhHS1YNpNBFEsdrGAE/89wv++k4BpTWNzBmbxvdOHces3FREJNLhKaX6kSYCh6ttDPDkf3fw1xWfU1LdyOTsZK48cQxnHZ2J26Wtg0o5gSYCBUBdYzPPri7kofe2sa2khqyUeBadkMvFM0bij9VnFCkVzTQRqHaaWwxvbCjigXcLWLl9P4mxbr46PZtLZ41i/LDESIenlAoDTQSqU2t3lvPw+9t49dO9NDa3MDMnlctmj+LMozOJdbsiHZ5Sqo9oIlDdKq1u4NlVhTz54Q52lNWSmuDlounZfHVaNkdm6lGCUoOdJgLVYy0thve2lvDEf7/grY37CLQYJo1IYt60bM6bPIKMxNhIh6iU6gVNBKpXSqobeOnj3Ty/ehef7qrAFSOcPD6DC6dm8aWjhuLz6glmpQaLrhJBVPwnNzU1UVhYSH19faRDiTqzU2H2aWk0NQ+htrGZ2sZmmmv28p9Ve4l1u4j3uojzxBDTw/sS4uLiyM7OxuPxhDlypVRPRUUiKCwsJDExkZycHL1RKsyMMdQ0NFNR10RFXROBlhaMCL44N8nxHhLjPLhiQv8GxhhKS0spLCwkNze3nyNXSnUmKhJBfX29JoF+IiL449z449yMSImjpvFAUqioa0JESPC6SIr3kBjnbnflkYiQlpZGcXFxBEuglOoorLeVisiZIrJJRLaKyI0h5k8Qkf+ISIOI/PgwP+twVle9ICL4Y91kpcRzVGYiR2T4Sfd7aWo27C6vY9PeKjbtrWJPRR01DQGMMfo7KTUAhe2IQERcwL3A6UAhsFJEXjTGrA9arAz4PnBBuOJQ/UNESIh1kxDrZniy1fFdVV2AyvomSqobKa5qwBUjJHjdVDcEKCiuJjc9QRODUgNAOI8IZgJbjTEFxphGYAlwfvACxph9xpiVQFMY4wi70tJSpkyZwpQpU8jMzCQrK6ttvLGxsct18/Pz+f73v9/tZxx//PF9Euvy5cs599xz+2RbXYl1u0hPjGVMhp+JwxMZneojOd5DfVMz5bVNnPr7d5jz67f46bMf88+1uyipbgh7TEqp0MJ5jiAL2Bk0XgjM6s2GRORq4GqAUaNGHX5kfSwtLY21a9cCcOutt+L3+/nxjw+0dAUCAdzu0F91Xl4eeXkhr+hq54MPPuiTWCPBFRNDss9Lss8LQFNpLHdeeDTvbSnh9XVF/CO/EIBxQ/3MyE1lZk4qM3NTGZESH8mwlXKMcCaCUMf8vbppwRizGFgM1n0EXS1720vrWL+7sjcf06mJI5L45VcmHdI6CxcuJDU1lTVr1jBt2jQuvvhirr/+eurq6oiPj+fhhx/myCOPZPny5dx11128/PLL3HrrrezYsYOCggJ27NjB9ddf33a04Pf7qa6uZvny5dx6662kp6fz2WefMX36dJ544glEhGXLlnHDDTeQnp7OtGnTKCgo4OWXX+40xrKyMhYtWkRBQQE+n4/Fixdz7LHH8s477/CDH/wAsJp8VqxYQXV1NRdffDGVlZUEAgHuv/9+TjzxxF59n25XDJfNGs1ls0bT3GL4bFcF739ewkfbynhp7W6e+nAHYD19bWaulRRm5KRyRIY2JSkVDuFMBIXAyKDxbGB3GD9vwNm8eTNvvPEGLpeLyspKVqxYgdvt5o033uBnP/sZzz333EHrbNy4kbfffpuqqiqOPPJIvv3tbx90zf2aNWtYt24dI0aMYM6cObz//vvk5eVxzTXXsGLFCnJzc5k/f3638f3yl79k6tSpLF26lLfeeosrrriCtWvXctddd3HvvfcyZ84cqquriYuLY/HixZxxxhn8/Oc/p7m5mdra2j75jlwxwuSRKUwemcJ35lod4m3cW8lH28pYub2Md7cU88KaXQCk+DxMzk5hykjrNXlkCqkJ3j6JQyknC2ciWAmME5FcYBdwCXBpGD8P4JD33MPpa1/7Gi6XdflkRUUFCxYsYMuWLYgITU2hT4ucc845xMbGEhsby9ChQykqKiI7O7vdMjNnzmybNmXKFLZv347f72fMmDFt1+fPnz+fxYsXdxnfe++915aMTj31VEpLS6moqGDOnDnccMMNXHbZZcybN4/s7GxmzJjBokWLaGpq4oILLmDKlCmH89V0yhUjTBqRzKQRyXxzTi7GGLaV1LByexlrdpSzdmc5f35rCy32ceHoNJ+VFLJTmDIqhYnDk4jzaGd5Sh2KsCUCY0xARK4DXgdcwEPGmHUicq09/y8ikgnkA0lAi4hcD0w0xvRt206EJCQktA3/4he/4JRTTuGFF15g+/btzJ07N+Q6sbEH+vJxuVwEAoEeLdObrkJCrSMi3HjjjZxzzjksW7aM2bNn88Ybb3DSSSexYsUKXnnlFS6//HJ+8pOfcMUVVxzyZx4qEWFMhp8xGX4unmGdH6ppCPDprgrW7ixn7Y5yPiwo459rrYNNV4wwNsPPxBFJTBqRZL0PTybZp3cyK9WZsN5QZoxZBizrMO0vQcN7sZqMol5FRQVZWVkAPPLII32+/QkTJlBQUMD27dvJycnh6aef7nadk046iSeffJJf/OIXLF++nPT0dJKSkvj888855phjOOaYY/jPf/7Dxo0biY+PJysri6uuuoqamhpWr17dL4kglIRYN7PHpDF7TFrbtL0V9azdWc663RWs213JB5+XtDUpgXW+YdKIJCaNSGbiiCQmZCaSlRJPTCd3QSvlJFFxZ/Fg8NOf/pQFCxbwhz/8gVNPPbXPtx8fH899993HmWeeSXp6OjNnzux2nVtvvZVvfvObHHvssfh8Ph599FEA7r77bt5++21cLhcTJ07krLPOYsmSJfzud7/D4/Hg9/t57LHH+rwMhyMzOY4zkzM58+jMtmkl1Q2s313Jut2VrNtdwfrdlfx7QxGtB0I+r4txQ/2MG5bI+GF+xg9LZPywRIYnx+lJaeUoUdH76IYNGzjqqKMiFNHAUV1djd/vxxjDd7/7XcaNG8cPf/jDSId1kEj+XjUNATburWTT3mo2F1WxZV8Vm4uqKa46cB9DYqybscP8jB+ayLhhfo7I8JObnkD2kHh9xrMatKK+91Fl+dvf/sajjz5KY2MjU6dO5Zprrol0SANOQqyb6aNTmT46td30/TWNbNlXzaaiKrYUVbG5qIo3NhTxdP6BW2E8LmFUqo/cdD9jMhIYk55AbnoCuRkJZPhj9ShCDVp6RKD63WD6vcpqGtlWUk1BcQ0FJTVsK65hW0kN20praAy0tC2XGOsmNyOB0WkJjEqNZ1Sqj5FDfIxM9TE8OU6PJFTE6RGBUr2UmuAlNeHgI4jmFqtjvW0lVmIoKK6moKSGj3eW8+qnewi0HNjBcscII1Ls5JDqY6SdKFqTRYrPo0cTKqI0ESjVC64YsSt1HyeNz2g3L9Dcwp6KenaW1bJzfy07ymrZUVbHzrJa/rVuL6U17fufSvC6GJ4Sz4iUeEYkxzEiJZ7hyXFkpcQz3B7WeyNUOGkiUKqPuV0xbUkilOqGAIX7a9lRaiWJ3eX17C6vY3dFHet3V4bsgC8twWslipQ4hifHk5UST2ZyHEMTYxmWFMewpDjivZosVO9oIlCqn/lj3UzITGJCZlLI+fVNzRRV1rOrvI49QUlid3k9BcU1vLelhJrG5oPWS4xzMyzpQHIYmhTLsMS4dsNDk2L16EIdRBNBH5g7dy433XQTZ5xxRtu0u+++m82bN3Pfffd1us5dd91FXl4eZ599Nk899RQpKSntlgnVk2lHS5cuZfz48UycOBGAW265hZNOOonTTjvtsMoU3Bme6l9xHhej06wTz6EYY6isD1BUWU9RZT37KhsoqrLfK+vZV9XAyu1l7KtsoLG55aD1k+M9DE2MJd0fS5rfS7o/lnT7Pa3dsBefV6sIJ9BfuQ/Mnz+fJUuWtEsErTdg9cSyZcu6X6gTS5cu5dxzz21LBL/61a96vS01OIgIyfEekuM9jB+W2OlyxhjKa5sOShKtyaOkuoF1uyspqWqgquHgrkzAuumuNVmkJbRPEun+WFITvKT4PKQmeBni8+rRxiAVfYng1Rth76d9u83MY+CsX3c6+6KLLuLmm2+moaGB2NhYtm/fzu7duznhhBP49re/zcqVK6mrq+Oiiy7itttuO2j9nJwc8vPzSU9P58477+Sxxx5j5MiRZGRkMH36dMC6R2Dx4sU0NjYyduxYHn/8cdauXcuLL77IO++8wx133MFzzz3H7bffzrnnnstFF13Em2++yY9//GMCgQAzZszg/vvvJzY2lpycHBYsWMBLL71EU1MTzzzzDBMmTOi0fJHqrlodHhFhSIKXIQleJmR2vWx9UzOlNY2UVjdQWt1Isf1eUt1AaXUDJdWNFO6vZe3OcspqGmjp5KrzeI+LIT6P9bm+A0kixeclNWj6EJ+XIQkehvi8+LwuvWoqwqIvEURAWloaM2fO5LXXXuP8889nyZIlXHzxxYgId955J6mpqTQ3N/OlL32JTz75hGOPPTbkdlatWsWSJUtYs2YNgUCAadOmtSWCefPmcdVVVwFw88038+CDD/K9732P8847r63iD1ZfX8/ChQt58803GT9+PFdccQX3338/119/PQDp6emsXr2a++67j7vuuosHHnig0/INhO6qVXjFeVxkpVgnobvT0mIor2uipLqBsppGymsbKatpYn9tI/trGtlf22RNq21kV3kdZTWNVNR1/hBCryuGIQkeUuK9JMW7SY73kBTvISnO03bkkxQfPOxuG473aBLpC9GXCLrYcw+n1uah1kTw0EMPAfCPf/yDxYsXEwgE2LNnD+vXr+80Ebz77rtceOGF+HzW1SbnnXde27zPPvuMm2++mfLycqqrq9s1Q4WyadMmcnNzGT9+PAALFizg3nvvbUsE8+bNA2D69Ok8//zzXW5rIHZXrSInJkbs+yt6/iyI5hZDRV1TUOJopLy2ibLaxrYEUlHXREVdE7vL69mwp4rKuqZOm6xaeVzSljASg5NF3IGEkhjnxh/rtt87jrv1Zj+iMRFEyAUXXMANN9zA6tWrqaurY9q0aWzbto277rqLlStXMmTIEBYuXEh9fX2X2+ls72bhwoUsXbqUyZMn88gjj7B8+fIut9PdHeOtXVl31tV1d9saaN1Vq4HN1YvkAVYCqaq3EkRlXaAtWVTa06zpB4Yr6prYWVbbNi3QWRtWkHiPC3+cm8TW5NCWKDztEkZinKdtOX+cNT3Ba83zxbrwumIG7dGJJoI+4vf7mTt3LosWLWp7OlhlZSUJCQkkJydTVFTEq6++2ulzCMDqFnrhwoXceOONBAIBXnrppbb+gqqqqhg+fDhNTU08+eSTbV1aJyYmUlVVddC2JkyYwPbt29m6dWvbOYWTTz65V2UbrN1Vq8HPFSOk+KxzDIfKGENdUzPV9QGqGgJU1Qeorg9Q3dBEZduw9aqqb7LmN1jTS6pq26ZXNwQ6PScSzB0j+LwuEmLd7d+9bnyxbhLsaQleV9u4z+smIbbDu9dKLAleN3Ge/kkumgj60Pz585k3bx5LliwBYPLkyUydOpVJkyYxZswY5syZ0+X6rc82njJlCqNHj253kvX2229n1qxZjB49mmOOOaat8r/kkku46qqr+NOf/sSzzz7btnxcXBwPP/wwX/va19pOFl977bW9Klc0dVetnENE8Hnd+Lxuhh7Gdowx1DY224nhQHJoTTC1DQFqGpupbQxQ02C/NzZb0xua2VtZT21jMzUNAeu9MUBPu3iLEdoSg8/r5rJZo7jyxDGHUZrQtNM51e/091JOZoyhvqmFmsYAtQ1WgmmXPDomFfu9uiHAaUcN44KpWb36XO10TimlBggRId7rsroE8Uc6GoueLldKKYeLmkQw2Jq4nEp/J6UGnqhIBHFxcZSWlmolM8AZYygtLSUuLi7SoSilgkTFOYLs7GwKCwspLi6OdCiqG3FxcWRnZ0c6DKVUkKhIBB6Ph9zc3EiHoZRSg1JUNA0ppZTqPU0ESinlcJoIlFLK4QbdncUiUgx80cvV04GSPgxnMNAyO4OW2RkOp8yjjTEZoWYMukRwOEQkv7NbrKOVltkZtMzOEK4ya9OQUko5nCYCpZRyOKclgsWRDiACtMzOoGV2hrCU2VHnCJRSSh3MaUcESimlOtBEoJRSDueYRCAiZ4rIJhHZKiI3Rjqe3hKRkSLytohsEJF1IvIDe3qqiPxbRLbY70OC1rnJLvcmETkjaPp0EfnUnvcnGeBP3hYRl4isEZGX7fGoLrOIpIjIsyKy0f69j3NAmX9o/11/JiJ/F5G4aCuziDwkIvtE5LOgaX1WRhGJFZGn7ekfikhOt0EZY6L+BbiAz4ExgBf4GJgY6bh6WZbhwDR7OBHYDEwEfgvcaE+/EfiNPTzRLm8skGt/Dy573kfAcYAArwJnRbp83ZT9BuAp4GV7PKrLDDwKXGkPe4GUaC4zkAVsA+Lt8X8AC6OtzMBJwDTgs6BpfVZG4DvAX+zhS4Cnu40p0l9KP33xxwGvB43fBNwU6bj6qGz/BE4HNgHD7WnDgU2hygq8bn8fw4GNQdPnA3+NdHm6KGc28CZwKgcSQdSWGUiyK0XpMD2ay5wF7ARSsXpGfhn4cjSWGcjpkAj6rIyty9jDbqw7kaWreJzSNNT6B9aq0J42qNmHfFOBD4Fhxpg9APb7UHuxzsqeZQ93nD5Q3Q38FGgJmhbNZR4DFAMP281hD4hIAlFcZmPMLuAuYAewB6gwxvyLKC5zkL4sY9s6xpgAUAGkdfXhTkkEodoHB/V1syLiB54DrjfGVHa1aIhppovpA46InAvsM8as6ukqIaYNqjJj7clNA+43xkwFarCaDDoz6Mtst4ufj9UEMgJIEJFvdLVKiGmDqsw90JsyHnL5nZIICoGRQePZwO4IxXLYRMSDlQSeNMY8b08uEpHh9vzhwD57emdlL7SHO04fiOYA54nIdmAJcKqIPEF0l7kQKDTGfGiPP4uVGKK5zKcB24wxxcaYJuB54Hiiu8yt+rKMbeuIiBtIBsq6+nCnJIKVwDgRyRURL9YJlBcjHFOv2FcGPAhsMMb8IWjWi8ACe3gB1rmD1umX2FcS5ALjgI/sw88qEZltb/OKoHUGFGPMTcaYbGNMDtZv95Yx5htEd5n3AjtF5Eh70peA9URxmbGahGaLiM+O9UvABqK7zK36sozB27oI6/+l6yOiSJ806ceTM2djXWHzOfDzSMdzGOU4Aesw7xNgrf06G6sN8E1gi/2eGrTOz+1ybyLo6gkgD/jMnncP3ZxQGggvYC4HThZHdZmBKUC+/VsvBYY4oMy3ARvteB/HulomqsoM/B3rHEgT1t77t/qyjEAc8AywFevKojHdxaRdTCillMM5pWlIKaVUJzQRKKWUw2kiUEoph9NEoJRSDqeJQCmlHE4TgXIsEam233NE5NI+3vbPOox/0JfbV6ovaSJQyuoA7JASgYi4ulmkXSIwxhx/iDEp1W80ESgFvwZOFJG1dn/4LhH5nYisFJFPROQaABGZK9azIJ4CPrWnLRWRVXYf+lfb034NxNvbe9Ke1nr0Ifa2P7P7kr84aNvL5cDzB54cSH3oq+jmjnQASg0ANwI/NsacC2BX6BXGmBkiEgu8LyL/spedCRxtjNlmjy8yxpSJSDywUkSeM8bcKCLXGWOmhPiseVh3DE8G0u11VtjzpgKTsPqMeR+rj6X3+rqwSnWkRwRKHezLwBUishari+80rD5ewOrnZVvQst8XkY+B/2J19DWOrp0A/N0Y02yMKQLeAWYEbbvQGNOC1XVITh+URalu6RGBUgcT4HvGmNfbTRSZi9UddPD4aVgPAakVkeVY/bx0t+3ONAQNN6P/n6qf6BGBUlCF9djPVq8D37a7+0ZExtsPhekoGdhvJ4EJwOygeU2t63ewArjYPg+RgfXYwo/6pBRK9ZLucShl9e4ZsJt4HgH+iNUss9o+YVsMXBBivdeAa0XkE6yeIf8bNG8x8ImIrDbGXBY0/QWsRw1+jNWL7E+NMXvtRKJURGjvo0op5XDaNKSUUg6niUAppRxOE4FSSjmcJgKllHI4TQRKKeVwmgiUUsrhNBEopZTD/X9kFpbWFV9hFAAAAABJRU5ErkJggg==\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "pyplot.figure()\n", + "pyplot.plot(range(10000), hist_train_loss, label=\"Training loss\")\n", + "pyplot.plot(range(10000), hist_val_loss, label=\"Validation loss\")\n", + "pyplot.xlabel(\"Iteration\")\n", + "pyplot.ylabel(\"Loss\")\n", + "pyplot.title(\"Training/validation losses history\")\n", + "pyplot.legend(loc=3);" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We can see that after the 1472nd iteration, the training loss kept going down, but the validation loss bounced back. It indicates that, after this point, the model started to overfit the training data." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### Note\n", + "* In a real-world application, rather than running with a fixed number of iterations, we usually monitor the validation loss or/and accuracy in real-time and stop the training immediately when overfitting happens. \n", + "* In addition to validation loss, the accuracy (e.g., F-score) against validation data is also a typical quantity to monitor. Keep in mind that sometimes real-time F-score gives a false impression. For example, suppose the probabilities predicted by the logistic model during the last three iterations are $0.1$, $0.2$, and $0.3$., so the predicted labels are $0$, $0$, and $0$ for the three iterations. Hence monitoring F-score (which relies on predicted labels) will give a false impression that the optimization has reached a plateau, though it hasn't." ] }, { @@ -997,22 +1025,22 @@ }, { "cell_type": "code", - "execution_count": 24, + "execution_count": 25, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Final precision: 88.6%\n", + "Final precision: 89.8%\n", "Final recall: 84.6%\n", - "Final F-score: 86.6%\n" + "Final F-score: 87.1%\n" ] } ], "source": [ "# final accuracy\n", - "pred_labels_test = classify(images_test, (w, b))\n", + "pred_labels_test = classify(images_test, (best_w, best_b))\n", "perf = performance(pred_labels_test, labels_test)\n", "\n", "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", @@ -1024,7 +1052,7 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Bravo! The precision improved from $60.2\\%$ to $88.6\\%$, which means when our model says a part has defects, there is an $88.6\\%$ chance it is indeed defective. The recall dropped from $100%$ to $84.6\\%$, which means our model misses about $15\\%$ of the defective parts. The F-score, representing overall accuracy, improved from $75.2\\%$ to $86.6\\%$." + "Bravo! The precision improved from $60.2\\%$ to $89.8\\%$, which means when our model says a part has defects, there is an $89.8\\%$ chance it is indeed defective. The recall dropped from $100%$ to $84.6\\%$, which means our model misses about $15\\%$ of the defective parts. The F-score, representing overall accuracy, improved from $75.2\\%$ to $87.1\\%$." ] }, { @@ -1033,7 +1061,7 @@ "source": [ "##### Final note\n", "\n", - "No model is perfect, and it's not realistic to expect $100\\%$ accuracy. How accurate a model should be to consider it a good model is problem-dependent. Still, an F-score of $86.6\\%$ does not seem very exciting. And it's not pleasant to see about $15\\%$ of defective products slip through and may be handed over to customers. \n", + "No model is perfect, and it's not realistic to expect $100\\%$ accuracy. How accurate a model should be to consider it a good model is application-dependent. Still, an F-score of $87.1\\%$ does not seem very exciting. And it's not pleasant to see about $15\\%$ of defective products slip through and may be handed over to customers. \n", "\n", "In a later lesson, we will improve the performance by replacing the multiple logistic regression model with something more interesting: a **neural network** model." ] @@ -1063,7 +1091,7 @@ }, { "cell_type": "code", - "execution_count": 25, + "execution_count": 26, "metadata": {}, "outputs": [ { @@ -1223,7 +1251,7 @@ "" ] }, - "execution_count": 25, + "execution_count": 26, "metadata": {}, "output_type": "execute_result" } From 527049f102c9eb6ec2bea74bd1b6ca6c2c2d78ad Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Tue, 10 Aug 2021 13:54:32 -0400 Subject: [PATCH 6/8] add the term "feature" to the appendix --- notebooks_en/Appendix.ipynb | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/notebooks_en/Appendix.ipynb b/notebooks_en/Appendix.ipynb index 4687010..59375ec 100644 --- a/notebooks_en/Appendix.ipynb +++ b/notebooks_en/Appendix.ipynb @@ -35,6 +35,16 @@ "Nevertheless, the graphical illustration makes it's easier to understand why the deep learning community names the math components with the following terms." ] }, + { + "cell_type": "markdown", + "id": "fde3fb39", + "metadata": {}, + "source": [ + "##### Features\n", + "\n", + "The elements in an input vector are called features. Recall in lesson 3, the vector $x$ of each sample has some characteristics of a car. It's why the elements in $x$ are called *features*." + ] + }, { "cell_type": "markdown", "id": "13d31e91", From 282b4d314c83c1c3e0bebf194be90583dd61e8c6 Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Tue, 10 Aug 2021 13:55:06 -0400 Subject: [PATCH 7/8] lesson 7: record loss history, use params at the best val loss --- .../7_The_First_Deep_Neural_Network.ipynb | 128 +++++++++++------- 1 file changed, 77 insertions(+), 51 deletions(-) diff --git a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb index 96fe342..006a487 100644 --- a/notebooks_en/7_The_First_Deep_Neural_Network.ipynb +++ b/notebooks_en/7_The_First_Deep_Neural_Network.ipynb @@ -52,7 +52,7 @@ "metadata": {}, "outputs": [], "source": [ - "numpy.random.seed(10000)" + "numpy.random.seed(0)" ] }, { @@ -654,15 +654,15 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "The optimization is the same as how we did it in lesson 5, except that now we have to update `W0`, `b0`, `W1`, and `b1` (in lesson 5, we only updated `W` and `b`). Also, we give you a proper step size (`lr`) directly." + "The optimization is the same as how we did it in lesson 5, except that now we have to update `W0`, `b0`, `W1`, and `b1` (in lesson 5, we only updated `W` and `b`). Also, we give you a proper step size (`lr`) directly.\n", + "\n", + "(The following code cell takes around 2.5 minutes to finish on a personal desktop.)" ] }, { "cell_type": "code", "execution_count": null, - "metadata": { - "scrolled": true - }, + "metadata": {}, "outputs": [], "source": [ "%%time\n", @@ -670,44 +670,47 @@ "# step size\n", "lr = 1e-1\n", "\n", - "# a variable for the change in validation loss\n", - "change = numpy.inf\n", - "\n", - "# a counter for optimization iterations\n", - "i = 0\n", + "# variables to store parameters at when the validation loss is at its lowest\n", + "best_val_loss = numpy.inf\n", + "best_iter = 0\n", + "best_W0 = numpy.zeros_like(W0)\n", + "best_b0 = numpy.zeros_like(b0)\n", + "best_W1 = numpy.zeros_like(W1)\n", + "best_b1 = 0.\n", "\n", - "# a variable to store the validation loss from the previous iteration\n", - "old_val_loss = 1e-15\n", + "# variables to store the history of training and validation loss\n", + "hist_train_loss = []\n", + "hist_val_loss = []\n", "\n", - "# keep running if:\n", - "# 1. we still see significant changes in validation loss\n", - "# 2. iteration counter < 10000\n", - "while change >= 1e-6 and i < 10000:\n", + "# run for 2000 iterations\n", + "for i in range(2000):\n", " \n", " # calculate gradients and use gradient descents\n", " grads = gradients(images_train, labels_train, (W0, b0, W1, b1))\n", - " W0 -= (grads[0] * lr)\n", - " b0 -= (grads[1] * lr)\n", - " W1 -= (grads[2] * lr)\n", - " b1 -= (grads[3] * lr)\n", + " W0 = W0 - grads[0] * lr\n", + " b0 = b0 - grads[1] * lr\n", + " W1 = W1 - grads[2] * lr\n", + " b1 = b1 - grads[3] * lr\n", " \n", - " # validation loss\n", - " val_loss = model_loss(images_val, labels_val, (W0, b0, W1, b1))\n", + " # save losses for future investigation\n", + " hist_train_loss.append(model_loss(images_train, labels_train, (W0, b0, W1, b1)))\n", + " hist_val_loss.append(model_loss(images_val, labels_val, (W0, b0, W1, b1)))\n", " \n", - " # calculate f-scores against the validation dataset\n", - " pred_labels_val = classify(images_val, (W0, b0, W1, b1), neural_network_model)\n", - " score = performance(pred_labels_val, labels_val)\n", - "\n", - " # calculate the chage in validation loss\n", - " change = numpy.abs((val_loss-old_val_loss)/old_val_loss)\n", - "\n", - " # update the counter and old_val_loss\n", - " i += 1\n", - " old_val_loss = val_loss\n", + " # if the validation loss is lower than the current best, save it\n", + " if hist_val_loss[-1] < best_val_loss:\n", + " best_val_loss = hist_val_loss[-1]\n", + " best_iter = i\n", + " best_W0 = W0\n", + " best_b0 = b0\n", + " best_W1 = W1\n", + " best_b1 = b1\n", " \n", - " # print the progress every 10 steps\n", - " if i % 10 == 0:\n", - " print(\"{}...\".format(i), end=\"\")" + " # print something every 100 iterations so we know it's not dead\n", + " if i % 100 == 0:\n", + " print(\"{}...\".format(i), end=\"\")\n", + "\n", + "# just to add a new line\n", + "print()" ] }, { @@ -716,25 +719,46 @@ "metadata": {}, "outputs": [], "source": [ - "val_loss_reg = regularized_loss(images_train, labels_train, (W0, b0, W1, b1))\n", + "print(\"The best validation loss:\", best_val_loss)\n", + "print(\"It happened at iteration {}\".format(best_iter))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The parameters that give us the best validation performance happened at the $456$th iteration. As we can see from the following figure, after the $456$th iteration, the validation loss started to go up, while the training loss kept going down." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "pyplot.figure()\n", + "pyplot.semilogy(range(2000), hist_train_loss, label=\"Training loss\")\n", + "pyplot.semilogy(range(2000), hist_val_loss, label=\"Validation loss\")\n", + "pyplot.xlabel(\"Iteration\")\n", + "pyplot.ylabel(\"Loss\")\n", + "pyplot.title(\"Training/validation losses history\")\n", + "pyplot.legend(loc=3);" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "##### Think about it\n", "\n", - "print(\"\")\n", - "print(\"\")\n", - "print(\"Upon optimization stopped:\")\n", - "print(\" Iterations:\", i)\n", - "print(\" Validation loss w/o regularization:\", val_loss)\n", - "print(\" Validation loss w/ regularization:\", val_loss_reg)\n", - "print(\" Validation precision:\", score[0])\n", - "print(\" Validation recall:\", score[1])\n", - "print(\" Validation F-score:\", score[2])\n", - "print(\" Change in validation loss:\", change)" + "* Why is there oscillation in the loss history?" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "Finally, let's check the final model performance:" + "Finally, let's check the model performance after optimization:" ] }, { @@ -744,7 +768,7 @@ "outputs": [], "source": [ "# final accuracy\n", - "pred_labels_test = classify(images_test, (W0, b0, W1, b1), neural_network_model)\n", + "pred_labels_test = classify(images_test, (best_W0, best_b0, best_W1, best_b1), neural_network_model)\n", "perf = performance(pred_labels_test, labels_test)\n", "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", "print(\"Final recall: {:.1f}%\".format(perf[1]*100))\n", @@ -755,16 +779,18 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Awesome! Compared to lesson 5, simply replacing the logistic regression model with a fully connected neural network model with $L=1$ improves the F-score from $86.6\\%$ to $92.7\\%$. The precision is now $91.8\\%$, meaning whenever our model predicts a casting part being defective, $91.8\\%$ chance it's correct. And the recall is now $93.4\\%$, which says that our model misses around $6.6\\%$ of defective parts." + "Awesome! Compared to lesson 5, simply replacing the logistic regression model with a fully connected neural network model with $L=1$ improves the F-score from $87.1\\%$ to $93.1\\%$. The precision is now $91.4\\%$, meaning whenever our model predicts a casting part being defective, $91.4\\%$ chance it's correct. And the recall is now $94.9\\%$, which says that our model misses around $5\\%$ of defective parts." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "Of course, the performance may still be unsatisfying. However, one benefit of using neural networks is that we can control the complexity of our model. For example, we may get better performance if we increase $L$ and the lengths of intermediate results $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. \n", + "The performance may still be unsatisfying. However, one benefit of using neural networks is that we can control the complexity of our model. For example, we may get better performance if we increase $L$ and the lengths of intermediate results $\\mathbf{z}^1$, $\\mathbf{z}^2$, $\\dots$, $\\mathbf{z}^L$. \n", + "\n", + "However, what we didn't address in this lesson is that increasing the complexity increases the difficulty of optimization. Optimization may need much more time to converge or may not even find a satisfying solution. So it's not like we can freely increase $L$ and the lengths of intermediate results to whatever values we like. One way to ease the optimization when having high $L$ and sizes is to use advanced optimization methods. In lesson 8, we will see more different optimization methods. \n", "\n", - "What we didn't address in this lesson, however, is that increasing the complexity increases the difficulty of optimization as well. Optimization may need much more time to converge or may not even find a satisfying solution. So it's not like we can freely increase $L$ and the lengths of intermediate results to whatever values we like. One way to ease the optimization when having high $L$ and lengths is to use advanced optimization methods. In lesson 8, we will see more different optimization methods. Another approach is to use other types of neural networks, such as a convolutional neural network, which is a more popular choice for computer vision applications. We will talk about this in a later lesson. " + "Another approach is to use other types of neural networks, such as a convolutional neural network, which is a more popular choice for computer vision applications. We encourage those who are interested in advanced neural networks to keep learning after this module. Though we won't be learning more advanced neural networks in this module, you get the basic idea of machine/deep learning: assuming a model and optimize the model parameters to fit existing data." ] }, { From acdafaf244201b7d4cb924e20d8ca4efd0c9d517 Mon Sep 17 00:00:00 2001 From: Pi-Yueh Chuang Date: Tue, 24 Aug 2021 15:47:48 -0400 Subject: [PATCH 8/8] add lesson 8 --- notebooks_en/8_More_about_optimization.ipynb | 978 +++++++++++++++++++ scripts/lesson_5_functions.py | 1 + scripts/lesson_7_functions.py | 100 ++ 3 files changed, 1079 insertions(+) create mode 100644 notebooks_en/8_More_about_optimization.ipynb create mode 100644 scripts/lesson_7_functions.py diff --git a/notebooks_en/8_More_about_optimization.ipynb b/notebooks_en/8_More_about_optimization.ipynb new file mode 100644 index 0000000..368759b --- /dev/null +++ b/notebooks_en/8_More_about_optimization.ipynb @@ -0,0 +1,978 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "bb6b004f", + "metadata": {}, + "source": [ + "# More about optimization" + ] + }, + { + "cell_type": "markdown", + "id": "96a88e50", + "metadata": {}, + "source": [ + "In previous lessons, we use *optimization*, *learning*, and *training* interchangeably. They all mean finding the best set of model parameters. However, optimization has a broader meaning -- any applications in which we want to find parameters that best meet our goals belong to optimization. It is itself a topic of study and research.\n", + "\n", + "Gradient descent is an entry-level optimization method. The first part of this lesson will introduce stochastic gradient descent, an optimization method derived from vanilla gradient descent. In the second half of this lesson, we will talk about using the optimizers from SciPy. SciPy is a well-known Python package for scientific\n", + "computing. It has several advanced optimizers that may be too complicated for us to write from scratch. Using third-party packages like SciPy can save us a lot of time.\n", + "\n", + "Our goal is to give you a taste of different optimizations and how to use third-party packages. We will refrain from digging too deep into the math.\n", + "\n", + "##### Warm-up exercise\n", + "\n", + "Consider the following problem: *a to-be-built rectangular fence will enclose $120$ square feet. The material cost for the long and short sides will be $\\$5$ and $\\$6$ per foot, respectively. What are the dimensions of the fence that will minimize cost?*\n", + "\n", + "Let the lengths of the long and short sides be $l$ and $s$. This is an optimization problem as we want to find a set of $l$ and $s$ that gives us the best cost $c = 2\\left(5l+6s\\right)$. Try to solve this problem with gradient descent. The solutions are $l=12$ and $s=10$. (Hint: substitute $s = 120l^{-1}$ into the cost first.)" + ] + }, + { + "cell_type": "markdown", + "id": "a2f2c099", + "metadata": {}, + "source": [ + "### Before we start" + ] + }, + { + "cell_type": "markdown", + "id": "eb69a275", + "metadata": {}, + "source": [ + "We again use the application of identifying defective metal-casting parts. Let's import the required packages and normalized data:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "7b6ca0b3", + "metadata": {}, + "outputs": [], + "source": [ + "from autograd import numpy\n", + "from autograd import grad\n", + "from matplotlib import pyplot" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6c61563d", + "metadata": {}, + "outputs": [], + "source": [ + "\"\"\"The following lines imports variables and functions from lesson 5 and 7. The\n", + "imported variables and functions include : images_train, images_val, images_test,\n", + "labels_train, labels_val, labels_test, neural_network_model, model_loss,\n", + "regularized_loss, and performance.\"\"\"\n", + "import sys\n", + "sys.path.insert(0, \"../scripts\")\n", + "from load_casting_data import *\n", + "from lesson_7_functions import *" + ] + }, + { + "cell_type": "markdown", + "id": "0a22e81b", + "metadata": {}, + "source": [ + "In addition, we define a function to return a new set of parameters with Xavier initialization. Every time we train the model with a new optimization method, we want to re-initialize the parameters to the same values to have a fair comparison between different methods. Note that we use a length-one NumPy array for $b^1$ (which is supposed to be a scalar) because then we can use many NumPy functions on it later." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "de891498", + "metadata": {}, + "outputs": [], + "source": [ + "def get_new_parameters(n0, n1):\n", + " \"\"\"Get a new set of parameters using a fixed seed and the Xavier initialization.\n", + " \n", + " Arguments\n", + " ---------\n", + " n0, n1 : int\n", + " The number of elements in vector x and the first intermediate vector z1.\n", + " \n", + " Returns\n", + " -------\n", + " A list of 4 elementsb\n", + " W0 : a 2D array of shape (n0, n1).\n", + " b0 : a 1D array of length n1.\n", + " W1 : a 1D array of length n1.\n", + " b1 : a 1D array of only one element (i.e., to mimic a scalar).\n", + " \"\"\"\n", + " numpy.random.seed(0) # fix the random seed to avoid \"true randomness\"\n", + " \n", + " scale = numpy.sqrt(6/(n0+n1))\n", + " W0 = numpy.random.uniform(-scale, scale, (n0, n1))\n", + " b0 = numpy.zeros(n1)\n", + "\n", + " scale = numpy.sqrt(6/(n1+1))\n", + " W1 = numpy.random.uniform(-scale, scale, n1)\n", + " b1 = numpy.zeros(1) # using a length-one vector as a scalar\n", + " \n", + " return [W0, b0, W1, b1]" + ] + }, + { + "cell_type": "markdown", + "id": "636db73b", + "metadata": {}, + "source": [ + "The lengths of each input vector $\\mathbf{x}$ and the intermediate vector $\\mathbf{z}^1$ are fixed throughout this lesson." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9e6002dc", + "metadata": {}, + "outputs": [], + "source": [ + "n0 = images_train.shape[1] # i.e., total num. of pixels per image\n", + "n1 = 64 # the length of z1 vector" + ] + }, + { + "cell_type": "markdown", + "id": "2926e745", + "metadata": {}, + "source": [ + "Don't forget the function that calculates the gradients for us: " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6fb9e003", + "metadata": {}, + "outputs": [], + "source": [ + "gradients = grad(regularized_loss, argnum=2)" + ] + }, + { + "cell_type": "markdown", + "id": "eb6671ef", + "metadata": {}, + "source": [ + "## Revisiting gradient descent" + ] + }, + { + "cell_type": "markdown", + "id": "f96c0067", + "metadata": {}, + "source": [ + "In gradient descent, the gradients are the derivatives of a total loss $L$ with respect to parameters. The total loss can be the sum or the mean of the losses from all training samples. For example, we use the mean when doing logistic regression over $N$ training samples:\n", + "\n", + "$$\n", + "L\\left(\\hat{\\mathbf{y}}, \\mathbf{y}\\right)\n", + "=\n", + "\\frac{1}{N} \\sum_{i=1}^{N} l\\left(\\hat{y}_i, y_i\\right)\n", + "=\n", + "\\frac{1}{N}\\sum_{i=1}^{N} −y_i\\log(\\hat{y}_i)-(1-y_i)\\log(1-\\hat{y}_i)\n", + "$$\n", + "\n", + "The gradients, in this case, are also the mean of the gradients calculated from the single-sample loss, $l\\left(\\hat{y}_i, y_i\\right)$:\n", + "\n", + "$$\n", + "\\frac{\\mathrm{d}~L}{\\mathrm{d}~W}\\left(\\hat{\\mathbf{y}}, \\mathbf{y}\\right) = \\frac{1}{N}\\sum_{i=1}^{N}\\frac{\\mathrm{d}~l}{\\mathrm{d}~W}\\left(\\hat{y}_i, y_i\\right)\n", + "$$\n", + "\n", + "And the gradient descent algorithm is roughly\n", + "\n", + "```\n", + "loop:\n", + " Y_hat 🠘 model(X, W)\n", + " L 🠘 loss(Y_hat, Y)\n", + " W 🠘 W - step_size x (dL/dW)\n", + " exit optimization if model performance is satisfying\n", + "```\n", + "\n", + "(Note we never implemented code to calculate `dL/dW` (i.e., $\\frac{1}{N}\\sum_{i=1}^{N}\\frac{\\mathrm{d}~l}{\\mathrm{d}~W}\\left(y_i, \\hat{y}_i\\right)$) because `autograd` did that for us.)" + ] + }, + { + "cell_type": "markdown", + "id": "ec92537a", + "metadata": {}, + "source": [ + "## Batched and stochastic gradient descent" + ] + }, + { + "cell_type": "markdown", + "id": "9e1e7510", + "metadata": {}, + "source": [ + "When we have more samples in the training dataset, an iteration of gradient descent becomes more computationally expensive, i.e., it needs more time and computer memory. This is when ***batched gradient descent*** comes into play." + ] + }, + { + "cell_type": "markdown", + "id": "8419ea89", + "metadata": {}, + "source": [ + "Batched gradient descent divides an entire training dataset into several batches. Next, it assumes the gradients calculated from the loss of each batch are about the same as the overall gradients:\n", + "\n", + "$$\n", + "\\frac{\\mathrm{d}~L}{\\mathrm{d}~W}\\left(\\hat{\\mathbf{y}}, \\mathbf{y}\\right)\n", + "= \\frac{1}{N}\\sum_{i=1}^{N}\\frac{\\mathrm{d}~l}{\\mathrm{d}~W}\\left(\\hat{y}_i, y_i\\right)\n", + "\\approx \\frac{1}{N_b}\\sum_{i=1}^{N_b}\\frac{\\mathrm{d}~l}{\\mathrm{d}~W}\\left(\\hat{y}_i, y_i\\right)\n", + "\\approx \\frac{1}{N_b}\\sum_{i=N_b+1}^{2N_b}\\frac{\\mathrm{d}~l}{\\mathrm{d}~W}\\left(\\hat{y}_i, y_i\n", + "\\right)\n", + "\\approx \\cdots\n", + "$$\n", + "\n", + "$N_b$ denotes the number of samples in each batch. Batches don't need to have the same number of samples. For example, the last batch can be smaller because $N$ may not be divisible by $N_b$.\n", + "\n", + "Some people use batched gradient descent and ***stochastic gradient descent*** interchangeably. Some people, on the other hand, specifically use stochastic gradient descent for the case $N_b=1$. \n", + "\n", + "The data distribution should be similar in each batch to make batch gradients close enough to overall gradients. For example, let's say we have 800 good casting parts and 200 defective casting parts in our training dataset. In that case, each batch should also have approximately 80%/20% of good/defective casting parts. We can achieve this by ***shuffling*** the training data before dividing them into batches." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "d39f6e42", + "metadata": {}, + "outputs": [], + "source": [ + "# fix the random seed to avoid \"true randomness\"\n", + "numpy.random.seed(12345)\n", + "\n", + "# create an array for original indices\n", + "order = numpy.arange(len(images_train))\n", + "\n", + "# shuffle the indices (`order` is re-ordered in-place)\n", + "numpy.random.shuffle(order)\n", + "\n", + "# get re-ordered training data\n", + "images_train = images_train[order]\n", + "labels_train = labels_train[order]" + ] + }, + { + "cell_type": "markdown", + "id": "3171d76d", + "metadata": {}, + "source": [ + "Note that in the above code cell, the function `numpy.random.shuffle` can shuffle `images_train` directly. However, we first created `new_order` instead because we need to re-order `labels_train` with the same order as `images_train`.\n", + "\n", + "We use $N_b=100$ here. The total number of samples in the training dataset is $782$, so we have $8$ batches (with the last batch only having $82$ samples)." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6b42d46d", + "metadata": {}, + "outputs": [], + "source": [ + "# define the batch size Nb\n", + "Nb = 100\n", + "\n", + "# define the start and end index of each batch\n", + "starts = [0, 100, 200, 300, 400, 500, 600, 700] # python starts from 0\n", + "ends = [100, 200, 300, 400, 500, 600, 700, 782] # python uses end+1 for slicing" + ] + }, + { + "cell_type": "markdown", + "id": "bacc31f0", + "metadata": {}, + "source": [ + "Let's examine if each batch has a similar ratio between good and defective casting parts:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "35161eda", + "metadata": {}, + "outputs": [], + "source": [ + "# the overall good-defective ratio in the training dataset\n", + "Ndef = numpy.count_nonzero(labels_train)\n", + "print(\"The overall good-defective ratio in images_train: {}\".format(\n", + " (len(images_train)-Ndef)/Ndef))\n", + "\n", + "for i, (bg, ed) in enumerate(zip(starts, ends)):\n", + " # calculate and print the ratio\n", + " Ndef = numpy.count_nonzero(labels_train[bg:ed])\n", + " print(\"Batch {} (samples {} to {}): the ratio is {}\".format(\n", + " i, bg, ed-1, (ed-bg-Ndef)/Ndef))" + ] + }, + { + "cell_type": "markdown", + "id": "67affb21", + "metadata": {}, + "source": [ + "There's no golden rule how close these numbers should be. We assume the good-defective ratios from the batches are close enough to the overall one for the sake of convenience." + ] + }, + { + "cell_type": "markdown", + "id": "cf295565", + "metadata": {}, + "source": [ + "When updating the parameters, we loop through batches, calculate the loss of the current batch, get batch gradients, and then update the parameters using the batch gradients:\n", + "\n", + "```\n", + "loop:\n", + " loop i = [0, 1, ..., number_of_batches-1]:\n", + " X_b 🠘 training data in the i-th batch\n", + " Y_b 🠘 labels in the i-th batch\n", + " Y_hat_b 🠘 model(X_b, W)\n", + " L_b 🠘 loss(Y_hat_b, Y_b)\n", + " W 🠘 W - step_size x (dL_b/dW)\n", + " exit optimization if model performance is satisfying\n", + "```" + ] + }, + { + "cell_type": "markdown", + "id": "bfc5b329", + "metadata": {}, + "source": [ + "The `L_b` in the pseudo-code denotes the loss of `i`-th batch, $L_b = \\frac{1}{N_b}\\sum_{j=iN_b}^{(i+1)N_b}l\\left(\\hat{y}_j, y_j\\right)$, and `dL_b/dW` denotes its gradients." + ] + }, + { + "cell_type": "markdown", + "id": "e840bc88", + "metadata": {}, + "source": [ + "Let's try the new method to optimize our neural network model:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "038a8169", + "metadata": {}, + "outputs": [], + "source": [ + "%%time\n", + "\n", + "# get a new set of parameters (params=[W0, b0, W1, b1])\n", + "params = get_new_parameters(n0, n1)\n", + "\n", + "# step size\n", + "lr = 1e-1\n", + "\n", + "# variables to store parameters at when the validation loss is at its lowest\n", + "best_val_loss = numpy.inf\n", + "best_iter = 0\n", + "best_params = None\n", + "\n", + "# variables to store the history of training and validation loss\n", + "hist_train_loss = []\n", + "hist_val_loss = []\n", + "\n", + "# a counter for the numbero of total iterations\n", + "i = 0\n", + "\n", + "# limit the total number of iterations to 2000\n", + "while i < 2000:\n", + " \n", + " # loop through each batch\n", + " for j, (bg, ed) in enumerate(zip(starts, ends)):\n", + " \n", + " # alias for data in this batch to shorten the code\n", + " x_b = images_train[bg:ed]\n", + " labels_b = labels_train[bg:ed]\n", + " \n", + " # calculate gradients and use gradient descents\n", + " grads = gradients(x_b, labels_b, params)\n", + " params = [p - lr * g for p, g in zip(params, grads)]\n", + " \n", + " # save losses for future investigation\n", + " hist_train_loss.append(model_loss(x_b, labels_b, params))\n", + " hist_val_loss.append(model_loss(images_val, labels_val, params))\n", + " \n", + " # if the validation loss is lower than the current best, save it\n", + " if hist_val_loss[-1] < best_val_loss:\n", + " best_val_loss = hist_val_loss[-1]\n", + " best_iter = i\n", + " best_params = params\n", + " \n", + " # update iteration counter\n", + " i = i + 1\n", + " \n", + " # print something every 100 iterations so we know it's not dead\n", + " if i % 100 == 0:\n", + " print(\"{}...\".format(i), end=\"\")\n", + "\n", + "# just to add a new line\n", + "print()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "4e68a769", + "metadata": {}, + "outputs": [], + "source": [ + "print(\"The best validation loss:\", best_val_loss)\n", + "print(\"It happened at iteration {}\".format(best_iter))" + ] + }, + { + "cell_type": "markdown", + "id": "e729593e", + "metadata": {}, + "source": [ + "The optimization took about $100$ seconds to finish $2000$ iterations on a personal desktop with 6 physical cores. Compared to the $2.5$ minutes in lesson 7, it is about $1.5\\texttt{x}$ faster. Moreover, we now get the best parameters earlier at the $390$th iteration. If we monitor and stop the optimization in real-time, it means the time-to-solution is $19.5$ seconds. ($100 \\times 390 \\div 2000 = 19.5$). The time-to-solution of the vanilla gradient descent in lesson 7 is $34.2$ seconds ($2.5 \\times 60 \\times 456 \\div 2000 = 34.2$). So using batched gradient descent improves the time-to-solution with a $1.75\\texttt{x}$ speedup!\n", + "\n", + "Just so that you know: it's also possible to further improve the performance by dividing validation data into batches." + ] + }, + { + "cell_type": "markdown", + "id": "1e87a110", + "metadata": {}, + "source": [ + "We examine the loss history in the following figures:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "50a27396", + "metadata": {}, + "outputs": [], + "source": [ + "plt, ax = pyplot.subplots(1, 2, figsize=(10, 5))\n", + "ax[0].plot(range(len(hist_train_loss)), hist_train_loss, label=\"Training loss\")\n", + "ax[0].plot(range(len(hist_val_loss)), hist_val_loss, label=\"Validation loss\")\n", + "ax[0].set_xlabel(\"Iteration\")\n", + "ax[0].set_ylabel(\"Loss\")\n", + "ax[0].set_title(\"Iteration 0 to 1999\")\n", + "ax[0].legend(loc=0)\n", + "ax[1].plot(range(100), hist_train_loss[:100], label=\"Training loss\")\n", + "ax[1].plot(range(100), hist_val_loss[:100], label=\"Validation loss\")\n", + "ax[1].set_xlabel(\"Iteration\")\n", + "ax[1].set_ylabel(\"Loss\")\n", + "ax[1].set_title(\"Iteration 0 to 100\")\n", + "ax[1].legend(loc=0)\n", + "plt.suptitle(\"Training and validation loss history\");" + ] + }, + { + "cell_type": "markdown", + "id": "358af9c8", + "metadata": {}, + "source": [ + "We have the complete loss history on the left. Oscillation happened from the beginning to the end. To have a clearer view of the oscillation, we plotted the first 100 iterations on the right. This oscillation is normal for batched gradient descent. In a nutshell, though we assume the characteristics of a batch (distributions, means, gradients, losses, etc.) are about the same as those of the whole training dataset, they are still different more or less. So using one batch at a time usually gives a bumpier loss history than using the entire dataset.\n", + "\n", + "The smaller the $N_b$ is, the more oscillation we get. The strictly stochastic gradient descent ($N_b=1$) hence usually shows a considerable fluctuation." + ] + }, + { + "cell_type": "markdown", + "id": "e0fa1f24", + "metadata": {}, + "source": [ + "The model's final performance:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6a8d74cc", + "metadata": {}, + "outputs": [], + "source": [ + "# final accuracy\n", + "pred_labels_test = classify(images_test, best_params, neural_network_model)\n", + "perf = performance(pred_labels_test, labels_test)\n", + "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", + "print(\"Final recall: {:.1f}%\".format(perf[1]*100))\n", + "print(\"Final F-score: {:.1f}%\".format(perf[2]*100))" + ] + }, + { + "cell_type": "markdown", + "id": "8fc01a53", + "metadata": {}, + "source": [ + "The performance is slightly different from what we got in lesson 7. Many factors contribute to the difference. The most obvious reason is that we were not using the best hyperparameters in both lessons." + ] + }, + { + "cell_type": "markdown", + "id": "8c979e70", + "metadata": {}, + "source": [ + "Though there are still many other gradient-descent-based optimization methods, we stop the discussion of gradient descent here. Among so many different optimization methods, gradient descent and its variants are the most important ones in modern deep learning. They require fewer computing resources, so optimizing models against millions or even billions of data becomes feasible. We encourage those who want to go deeper with deep learning to check reference [1]." + ] + }, + { + "cell_type": "markdown", + "id": "419d711a", + "metadata": {}, + "source": [ + "## Using advanced optimizers from SciPy" + ] + }, + { + "cell_type": "markdown", + "id": "a1496c17", + "metadata": {}, + "source": [ + "We have been implementing gradient descent from scratch in all lessons so far. Gradient descent is easy to implement and powerful in many use cases. However, sometimes we may want to use more advanced methods, but implementing advanced methods may be challenging. For example, [optimizers based on Newton's method](https://en.wikipedia.org/wiki/Quasi-Newton_method) are commonly used for small-size problems but are complicated. It may be more feasible to use third-party packages in such a situation instead of writing the code from scratch.\n", + "\n", + "We want to show you how to use the optimizers from a well-known Python package called SciPy. We will use the [L-BFGS-B optimizer](https://en.wikipedia.org/wiki/Limited-memory_BFGS) to optimize our neural network. L-BFGS-B is based on Newton's method. The primary goal is to show how to use SciPy, and L-BFGS-B serves as an example, so we'll skip the math.\n", + "\n", + "SciPy provides a uniform interface for all optimization methods through [`scipy.optimize.minimize`](https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.minimize.html). In this lesson, we will use this function with\n", + "\n", + "```\n", + "results = scipy.optimize.minimize(\n", + " fun = ,\n", + " x0 = ,\n", + " args = ,\n", + " method = ,\n", + " jac = \n", + ")\n", + "```\n", + "\n", + "Let's take a look at the arguments:\n", + "\n", + "##### 1. $\\texttt{fun}$\n", + " \n", + "`fun` is a function that returns both the loss and gradients. This function must have the model parameters as its first argument. All other arguments should be after model parameters. For example, we usually use `params` to represent model parameters in our lessons and use `x` and `y` to represent input data and labels. So if the function name is `loss_and_grad`, then the definition of the function should be something like:\n", + "```python\n", + "def loss_and_grad(params, x, y):\n", + " ...\n", + " return loss, grads\n", + "```\n", + "\n", + "And then `fun=loss_and_grad`.\n", + "\n", + "SciPy expects all parameters to be packed into a flattened 1D array. Previously, our `params` is a `list` of several `numpy.ndarray`. Now we need to flatten each `numpy.ndarray` and concatenate them into a single vector.\n", + "\n", + "##### 2. $\\texttt{x0}$\n", + "\n", + "SciPy uses different notation. It uses `x` to denote model parameters while we use `params`. `x0` represents the initial values of model parameters.\n", + "\n", + "##### 3. $\\texttt{args}$\n", + "\n", + "All arguments except for the model parameters should be provided through `args`. For example, `loss_and_grad` requires three arguments: model parameters, input data, and labels. During training, input data and labels are `images_train` and `labels_train`, respectively. So `args=(images_train, labels_train)`.\n", + "\n", + "##### 4. $\\texttt{method}$\n", + "\n", + "This is a string indicating what optimization method we want to use. [This SciPy tutorial](https://docs.scipy.org/doc/scipy/reference/tutorial/optimize.html#unconstrained-minimization-of-multivariate-scalar-functions-minimize) lists the methods allowed in SciPy. Here we use `mothod='L-BFGS-B'`.\n", + " \n", + "##### 5. $\\texttt{jac}$\n", + "\n", + "We use `jac=True` to let SciPy know the function provided in `fun` also returns gradients. (`fun` may return the loss only in some situations, and SciPy will then use other approaches to calculate gradients.) Since we already use `autograd` to get gradients, we don't need SciPy to calculate the gradients for us.\n", + "\n", + "##### 6. $\\texttt{callback}$\n", + "\n", + "If we provide a function to `callback`, SciPy will call this function after each optimization iteration. The `callback` function takes only one argument: the model parameters at the current optimization iteration. We can then do things like calculating the current validation loss or printing iteration information." + ] + }, + { + "cell_type": "markdown", + "id": "cfdaaa8b", + "metadata": {}, + "source": [ + "Now let's see how to use SciPy. First, all optimizers are under submodule `scipy.optimize`. We import it first." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "98941e50", + "metadata": {}, + "outputs": [], + "source": [ + "import scipy.optimize" + ] + }, + { + "cell_type": "markdown", + "id": "1bbeed1e", + "metadata": {}, + "source": [ + "As mentioned, SciPy needs all model parameters to be flattened and packed into a 1D array. We define `join_parameters` to achieve it and `split_parameters` to recover the 1D array back to a list of `numpy.ndarray`s. `split_parameters` takes in three arguments: model parameters, input data sample size, and the length of the intermediate vector. We need the sizes to know the shapes of $W^0$, $\\mathbf{b}^0$, $W^1$, and $\\mathbf{b}^1$." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f14a5e2d", + "metadata": {}, + "outputs": [], + "source": [ + "def join_parameters(params):\n", + " \"\"\"Flatten and concatenate several ndarrays into one single vector.\n", + " \n", + " Arguments\n", + " ---------\n", + " params : a list of numpy.ndarray\n", + " For example, in our neural network, params=[W0, b0, W1, b1].\n", + " \n", + " Returns\n", + " -------\n", + " params : a 1D numpy.ndarray\n", + " \"\"\"\n", + " params = numpy.concatenate([p.flatten() for p in params])\n", + " return params" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "dec20312", + "metadata": {}, + "outputs": [], + "source": [ + "def split_parameters(params, n0, n1):\n", + " \"\"\"Split an 1D array or parameters into W0, b0, W1, b1.\n", + " \n", + " Arguments\n", + " ---------\n", + " params : 1D numpy.ndarray\n", + " The parameters of a neural network packed in an 1D array.\n", + " n0 : integer\n", + " The number of element in the input vector x.\n", + " n1 : integer\n", + " The number of element in the intermediate vector z1.\n", + " \n", + " Returns\n", + " -------\n", + " A list of 4 elements\n", + " W0 : a 2D array of shape (n0, n1).\n", + " b0 : a 1D array of length n1.\n", + " W1 : a 1D array of length n1.\n", + " b1 : a scalar.\n", + " \"\"\"\n", + " \n", + " W0 = params[:n0*n1].reshape((n0, n1))\n", + " b0 = params[n0*n1:(n0+1)*n1].reshape((n1,))\n", + " W1 = params[(n0+1)*n1:(n0+2)*n1].reshape((n1,))\n", + " b1 = params[-1]\n", + " return [W0, b0, W1, b1]" + ] + }, + { + "cell_type": "markdown", + "id": "3f8e68f6", + "metadata": {}, + "source": [ + "Next is what we will provide to `fun`. It is a function taking five arguments: model parameters `params`, input data `x`, labels `y`, size of each input data sample `n0`, and the length of intermediate vector `n1`. The function returns both the loss and gradients. Gradients also need to be flattened and packed into a 1D array." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8cabaa27", + "metadata": {}, + "outputs": [], + "source": [ + "def loss_and_grad(params, x, y, n0, n1):\n", + " \"\"\"A function that returns loss and gradients for using SciPy's optimizers.\n", + " \n", + " Arguments\n", + " ---------\n", + " params : 1D numpy.ndarray\n", + " The parameters of a neural network packed in an 1D array.\n", + " x : numpy.ndarray\n", + " The input data. If a 1D array, it represents the features of one sample. If a\n", + " 2D array, the first dimension represents the number of samples, and the second\n", + " dimension represents the features.\n", + " y : numpy.ndarray\n", + " The output of the model. If x is a 1D array, y is a scalar. If x is a 2D array,\n", + " y is a 1D array.\n", + " n0 : integer\n", + " The number of element in the input vector x.\n", + " n1 : integer\n", + " The number of element in the intermediate vector z1.\n", + " \n", + " Returns\n", + " -------\n", + " loss : a scalar\n", + " The total loss (regularized).\n", + " grads : a 1D numpy.ndarray\n", + " The gradients of w.r.t. parameters packed into a single vector.\n", + " \"\"\"\n", + " # split and recover the shape of parameters\n", + " params = split_parameters(params, n0, n1)\n", + " \n", + " # get the regularized loss (params now is [W0, b0, W1, b1])\n", + " loss = regularized_loss(x, y, params)\n", + " \n", + " # get the gradients (grads is now [dW0, db0, dW1, db1])\n", + " grads = gradients(x, y, params)\n", + " \n", + " # concatenate all gradients into a 1D array\n", + " grads = join_parameters(grads)\n", + " \n", + " return loss, grads" + ] + }, + { + "cell_type": "markdown", + "id": "3a3dacd7", + "metadata": {}, + "source": [ + "We want to calculate the validation loss at the end of every iteration and record the best model parameters (the ones that give the lowest validation loss). We define the function `best_param_recorder` for this purpose:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "113e00e1", + "metadata": {}, + "outputs": [], + "source": [ + "def best_param_recorder(params):\n", + " \"\"\"A callback function to record the parameters that give the best validation loss.\n", + " \n", + " Arguments\n", + " ---------\n", + " params : 1D numpy.ndarray\n", + " The parameters of a neural network packed in an 1D array.\n", + " \n", + " Returns\n", + " -------\n", + " A boolean. This function always return False to let SciPy optimizers know they can\n", + " continue the optimization.\n", + " \n", + " Notes\n", + " -----\n", + " Variables `best_val_loss` and `best_params` should already exist outside this\n", + " function when this function is called.\n", + " \"\"\"\n", + " \n", + " # `global` indicates these variables will be the same as those outside the function\n", + " global best_val_loss, best_params\n", + " \n", + " # split and recover the shape of parameters\n", + " params = split_parameters(params, n0, n1)\n", + "\n", + " # calculate the validation loss\n", + " val_loss = model_loss(images_val, labels_val, params)\n", + "\n", + " # if the new validation loss is smaller than the current best, save it\n", + " if val_loss <= best_val_loss:\n", + " best_val_loss = val_loss\n", + " best_params = params.copy()\n", + "\n", + " return False" + ] + }, + { + "cell_type": "markdown", + "id": "a6d56f39", + "metadata": {}, + "source": [ + "`best_param_recorder` takes in only model parameters and can only return a boolean. This limitation comes from SciPy. So in order to record the best parameters, we use a keyword `global` in the function. `global` notifies the function that the subsequent variables will be obtained and synchronized with the same variables outside the function, i.e., in the global scope. In this way, we don't have to return the best parameters using `return` but are still able to access them from outside the function. " + ] + }, + { + "cell_type": "markdown", + "id": "2d0c3d00", + "metadata": {}, + "source": [ + "In other words, we need to decalre and initialize those global variables first before using the function `best_param_recorder`:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "df951612", + "metadata": {}, + "outputs": [], + "source": [ + "# we need these variables to be decalred outside the function `best_param_recorder`\n", + "best_val_loss = numpy.inf\n", + "best_params = None" + ] + }, + { + "cell_type": "markdown", + "id": "38f06fd9", + "metadata": {}, + "source": [ + "Finally, we can use `scipy.optimize.minimize` to optimize the model:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ce58f241", + "metadata": {}, + "outputs": [], + "source": [ + "%%time\n", + "\n", + "# get a new set of parameters\n", + "params = get_new_parameters(n0, n1) # params=[W0, b0, W1, b1]\n", + "params = join_parameters(params) # flattened & concatenated\n", + "\n", + "# optimize the model using SciPy\n", + "results = scipy.optimize.minimize(\n", + " loss_and_grad, params, args=(images_train, labels_train, n0, n1),\n", + " method='L-BFGS-B', jac=True, tol=1e-9, callback=best_param_recorder)" + ] + }, + { + "cell_type": "markdown", + "id": "251d826b", + "metadata": {}, + "source": [ + "The returned object `results` contains several information. We can access them through a dot `.`. For example, the number of iterations is `results.nit`:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "1d381a3e", + "metadata": {}, + "outputs": [], + "source": [ + "print(\"Number of iterations: {}\".format(results.nit))" + ] + }, + { + "cell_type": "markdown", + "id": "eed5afa7", + "metadata": {}, + "source": [ + "And the final model parameters are in `results.x` (recall that SciPy uses `x` to denote model parameters). [This webpage](https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.OptimizeResult.html#scipy.optimize.OptimizeResult) shows other information carried by `results`. Let's see the performance against the test dataset using the model parameters from the last iteration:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "346a3d36", + "metadata": {}, + "outputs": [], + "source": [ + "# final accuracy using the parameters from the last iteration\n", + "params = split_parameters(results.x, n0, n1)\n", + "pred_labels_test = classify(images_test, params, neural_network_model)\n", + "perf = performance(pred_labels_test, labels_test)\n", + "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", + "print(\"Final recall: {:.1f}%\".format(perf[1]*100))\n", + "print(\"Final F-score: {:.1f}%\".format(perf[2]*100))" + ] + }, + { + "cell_type": "markdown", + "id": "f877bbfd", + "metadata": {}, + "source": [ + "It doesn't look as nice as what we got previously using gradient descent and the batched version. However, remember these numbers are obtained using the model parameters from the last iteration. They are not the parameters that give the lowest validation. Let's calculate the test performance again using the best parameters (i.e., `best_params`):" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "16617ec6", + "metadata": {}, + "outputs": [], + "source": [ + "# final accuracy using the best parameters\n", + "pred_labels_test = classify(images_test, best_params, neural_network_model)\n", + "perf = performance(pred_labels_test, labels_test)\n", + "print(\"Final precision: {:.1f}%\".format(perf[0]*100))\n", + "print(\"Final recall: {:.1f}%\".format(perf[1]*100))\n", + "print(\"Final F-score: {:.1f}%\".format(perf[2]*100))" + ] + }, + { + "cell_type": "markdown", + "id": "ae8ea7c3", + "metadata": {}, + "source": [ + "These numbers are different from what we got from gradient descent. It's possible to tune the L-BFGS-B optimizer to get closer numbers. We can find all tunable options for L-BFGS-B optimizer [here](https://docs.scipy.org/doc/scipy/reference/optimize.minimize-lbfgsb.html). We will stop here as our primary goal is to show you how to use SciPy's optimizers." + ] + }, + { + "cell_type": "markdown", + "id": "b500410e", + "metadata": {}, + "source": [ + "## Challenge" + ] + }, + { + "cell_type": "markdown", + "id": "335c1c7c", + "metadata": {}, + "source": [ + "Recall the 120-square-feet fence we wanted to build at the beginning of this notebook. Can you write code to solve the same problem using SciPy's optimizers?" + ] + }, + { + "cell_type": "markdown", + "id": "084352c0", + "metadata": {}, + "source": [ + "## What we've learned" + ] + }, + { + "cell_type": "markdown", + "id": "f95e6918", + "metadata": {}, + "source": [ + "1. Using batched gradient descent to accelerate the optimization.\n", + "2. Using third-party libaraies, like SciPy, to optimize a model with advanced optimizers." + ] + }, + { + "cell_type": "markdown", + "id": "772c6efd", + "metadata": {}, + "source": [ + "## Reference\n", + "1. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Chapter 8. Optimization for Training Deep Models. In Deep Learning (pp. 271–325). MIT Press. http://www.deeplearningbook.org" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "0be9b20d", + "metadata": {}, + "outputs": [], + "source": [ + "# Execute this cell to load the notebook's style sheet, then ignore it\n", + "from IPython.core.display import HTML\n", + "css_file = '../style/custom.css'\n", + "HTML(open(css_file, \"r\").read())" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.6" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/scripts/lesson_5_functions.py b/scripts/lesson_5_functions.py index 33611e0..12dad63 100644 --- a/scripts/lesson_5_functions.py +++ b/scripts/lesson_5_functions.py @@ -36,6 +36,7 @@ def logistic(x): The function does not restrict the shape of the input array. The output has the same shape as the input. """ + x = numpy.clip(x, -300., 300.) return 1. / (1. + numpy.exp(-x)) diff --git a/scripts/lesson_7_functions.py b/scripts/lesson_7_functions.py new file mode 100644 index 0000000..ce9ffa8 --- /dev/null +++ b/scripts/lesson_7_functions.py @@ -0,0 +1,100 @@ +#! /usr/bin/env python3 +# -*- coding: utf-8 -*- +# vim:fenc=utf-8 + +"""The functions used in lesson 7. + +Usage: + from lesson_7_functions import * + +This will make the following functions available: + - classify + - performance + - neural_network_model + - model_loss +""" +from autograd import numpy +from lesson_5_functions import logistic, classify, performance + + +# doing `from lesson_y_functions import *` will import these objects +__all__ = ["neural_network_model", "model_loss", "regularized_loss", "classify", "performance"] + + +def neural_network_model(x, params): + """A fully-connected neural network with L=1. + + Arguments + --------- + x : numpy.ndarray + The input of the model. It's shape should be (n_images, n_total_pixels). + params : a tuple/list of four elements + - The first element is W0, a 2D array with shape (n_total_pixels, n_z1). + - The second elenment is b0, an 1D array with length n_z1. + - The third element is W1, an 1D array with length n_z1. + - The fourth element is b1, a scalar. + + Returns + ------- + yhat : numpy.ndarray + The predicted values obtained from the model. It's an 1D array with + length n_images. + """ + z1 = logistic(numpy.dot(x, params[0])+params[1]) + yhat = logistic(numpy.dot(z1, params[2])+params[3]) + return yhat + + +def model_loss(x, true_labels, params): + """Calculate the predictions and the loss w.r.t. the true values. + + Arguments + --------- + x : numpy.ndarray + The input of the model. The shape should be (n_images, n_total_pixels). + true_labels : numpy.ndarray + The true labels of the input images. Should be 1D and have length of + n_images. + params : a tuple/list of two elements + - The first element is W0, a 2D array with shape (n_total_pixels, n_z1). + - The second elenment is b0, an 1D array with length n_z1. + - The third element is W1, an 1D array with length n_z1. + - The fourth element is b1, a scalar. + + Returns + ------- + loss : a scalar + The summed loss. + """ + pred = neural_network_model(x, params) + + n_images = x.shape[0] + + # major loss + loss = - ( + numpy.dot(true_labels, numpy.log(pred+1e-15)) + + numpy.dot(1.-true_labels, numpy.log(1.-pred+1e-15)) + ) / n_images + + return loss + + +def regularized_loss(x, true_labels, params, _lambda=1.): + """Return the loss with regularization. + + Arguments + --------- + x, true_labels, params : + Parameters for function `model_loss`. + _lambda : float + The weight of the regularization term. Default: 0.01 + + Returns + ------- + loss : a scalar + The summed loss. + """ + loss = model_loss(x, true_labels, params) + Nw = params[0].shape[0] * params[0].shape[1] + params[2].size + reg = ((params[0]**2).sum() + (params[2]**2).sum()) / Nw + return loss + _lambda * reg