machinelearning_notebook/5_nn/2-mlp_bp.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 多层神经网络\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 神经元\n",
    "\n",
    "神经元和感知器本质上是一样的，只不过我们说感知器的时候，它的激活函数是阶跃函数；而当我们说神经元时，激活函数往往选择为sigmoid函数或tanh函数。如下图所示：\n",
    "\n",
    "![neuron](images/neuron.gif)\n",
    "\n",
    "计算一个神经元的输出的方法和计算一个感知器的输出是一样的。假设神经元的输入是向量$\\vec{x}$，权重向量是$\\vec{w}$(偏置项是$w_0$)，激活函数是sigmoid函数，则其输出y：\n",
    "$$\n",
    "y = sigmod(\\vec{w}^T \\cdot \\vec{x})\n",
    "$$\n",
    "\n",
    "sigmoid函数的定义如下：\n",
    "$$\n",
    "sigmod(x) = \\frac{1}{1+e^{-x}}\n",
    "$$\n",
    "将其带入前面的式子，得到\n",
    "$$\n",
    "y = \\frac{1}{1+e^{-\\vec{w}^T \\cdot \\vec{x}}}\n",
    "$$\n",
    "\n",
    "sigmoid函数是一个非线性函数，值域是(0,1)。函数图像如下图所示\n",
    "\n",
    "![sigmod_function](images/sigmod.jpg)\n",
    "\n",
    "sigmoid函数的导数是：\n",
    "\\begin{eqnarray}\n",
    "y & = & sigmod(x) \\tag{1} \\\\\n",
    "y' & = & y(1-y)\n",
    "\\end{eqnarray}\n",
    "\n",
    "可以看到，sigmoid函数的导数非常有趣，它可以用sigmoid函数自身来表示。这样，一旦计算出sigmoid函数的值，计算它的导数的值就非常方便。\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 神经网络是啥?\n",
    "\n",
    "![nn1](images/nn1.jpeg)\n",
    "\n",
    "神经网络其实就是按照一定规则连接起来的多个神经元。上图展示了一个全连接(full connected, FC)神经网络，通过观察上面的图，我们可以发现它的规则包括：\n",
    "\n",
    "* 神经元按照层来布局。\n",
    "    - 最左边的层叫做输入层，负责接收输入数据；\n",
    "    - 最右边的层叫输出层，我们可以从这层获取神经网络输出数据;\n",
    "    - 输入层和输出层之间的层叫做隐藏层，因为它们对于外部来说是不可见的。\n",
    "* 同一层的神经元之间没有连接。\n",
    "* 第N层的每个神经元和第N-1层的所有神经元相连(这就是full connected的含义)，第N-1层神经元的输出就是第N层神经元的输入。\n",
    "* 每个连接都有一个权值。\n",
    "\n",
    "上面这些规则定义了全连接神经网络的结构。事实上还存在很多其它结构的神经网络，比如卷积神经网络(CNN)、循环神经网络(RNN)，他们都具有不同的连接规则。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 计算神经网络的输出\n",
    "\n",
    "神经网络实际上就是一个输入向量$\\vec{x}$到输出向量$\\vec{y}$的函数，即：\n",
    "\n",
    "$$\n",
    "\\vec{y} = f_{network}(\\vec{x})\n",
    "$$\n",
    "根据输入计算神经网络的输出，需要首先将输入向量$\\vec{x}$的每个元素的值$x_i$赋给神经网络的输入层的对应神经元，然后根据式1依次向前计算每一层的每个神经元的值，直到最后一层输出层的所有神经元的值计算完毕。最后，将输出层每个神经元的值串在一起就得到了输出向量$\\vec{y}$。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "接下来举一个例子来说明这个过程，我们先给神经网络的每个单元写上编号。\n",
    "\n",
    "![nn2](images/nn2.png)\n",
    "\n",
    "* 输入层有三个节点，我们将其依次编号为1、2、3；\n",
    "* 隐藏层的4个节点，编号依次为4、5、6、7；\n",
    "* 最后输出层的两个节点编号为8、9。\n",
    "\n",
    "因为我们这个神经网络是全连接网络，所以可以看到每个节点都和上一层的所有节点有连接。比如，我们可以看到隐藏层的节点4，它和输入层的三个节点1、2、3之间都有连接，其连接上的权重分别为$w_{41}$,$w_{42}$,$w_{43}$。那么，我们怎样计算节点4的输出值$a_4$呢？\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "为了计算节点4的输出值，我们必须先得到其所有上游节点（也就是节点1、2、3）的输出值。节点1、2、3是输入层的节点，所以，他们的输出值就是输入向量$\\vec{x}$本身。按照上图画出的对应关系，可以看到节点1、2、3的输出值分别是$x_1$,$x_2$,$x_3$。我们要求输入向量的维度和输入层神经元个数相同，而输入向量的某个元素对应到哪个输入节点是可以自由决定的。\n",
    "\n",
    "一旦我们有了节点1、2、3的输出值，我们就可以根据式1计算节点4的输出值$a_4$：\n",
    "\n",
    "![eqn_3_4](images/eqn_3_4.png)\n",
    "\n",
    "上式的$w_{4b}$是节点4的偏置项，图中没有画出来。而$w_{41}$,$w_{42}$,$w_{43}$分别为节点1、2、3到节点4连接的权重，在给权重$w_{ji}$编号时，我们把目标节点的编号$j$放在前面，把源节点的编号$i$放在后面。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "同样，我们可以继续计算出节点5、6、7的输出值$a_5$,$a_6$,$a_7$。这样，隐藏层的4个节点的输出值就计算完成了，我们就可以接着计算输出层的节点8的输出值$y_1$：\n",
    "\n",
    "![eqn_5_6](images/eqn_5_6.png)\n",
    "\n",
    "同理，我们还可以计算出$y_2$的值。这样输出层所有节点的输出值计算完毕，我们就得到了在输入向量$\\vec{x} = (x_1, x_2, x_3)^T$时，神经网络的输出向量$\\vec{y} = (y_1, y_2)^T$。这里我们也看到，输出向量的维度和输出层神经元个数相同。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 神经网络的矩阵表示\n",
    "\n",
    "神经网络的计算如果用矩阵来表示会很方便，此外可以用优化加速算法提高计算速度。\n",
    "\n",
    "我们先来看看隐藏层的矩阵表示，隐藏层4个节点的计算依次排列出来：\n",
    "\n",
    "![eqn_hidden_units](images/eqn_hidden_units.png)\n",
    "\n",
    "接着，定义网络的输入向量$\\vec{x}$和隐藏层每个节点的权重向量$\\vec{w}$。令\n",
    "\n",
    "![eqn_7_12](images/eqn_7_12.png)\n",
    "\n",
    "代入到前面的一组式子，得到：\n",
    "\n",
    "![eqn_13_16](images/eqn_13_16.png)\n",
    "\n",
    "现在，我们把上述计算$a_4$, $a_5$,$a_6$,$a_7$的四个式子写到一个矩阵里面，每个式子作为矩阵的一行，就可以利用矩阵来表示它们的计算了。令\n",
    "\n",
    "![eqn_matrix1](images/eqn_matrix1.png)\n",
    "\n",
    "带入前面的一组式子，得到\n",
    "\n",
    "![formular_2](images/formular_2.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在式2中，$f$是激活函数，在本例中是$sigmod$函数；$W$是某一层的权重矩阵；$\\vec{x}$是某层的输入向量；$\\vec{a}$是某层的输出向量。式2说明神经网络的每一层的作用实际上就是先将输入向量左乘一个数组进行线性变换，得到一个新的向量，然后再对这个向量逐元素应用一个激活函数。\n",
    "\n",
    "每一层的算法都是一样的。比如，对于包含一个输入层，一个输出层和三个隐藏层的神经网络，我们假设其权重矩阵分别为$W_1$,$W_2$,$W_3$,$W_4$，每个隐藏层的输出分别是$\\vec{a}_1$,$\\vec{a}_2$,$\\vec{a}_3$，神经网络的输入为$\\vec{x}$，神经网络的输出为$\\vec{y}$，如下图所示：\n",
    "\n",
    "![nn_parameters_demo](images/nn_parameters_demo.png)\n",
    "\n",
    "则每一层的输出向量的计算可以表示为：\n",
    "\n",
    "![eqn_17_20](images/eqn_17_20.png)\n",
    "\n",
    "\n",
    "这就是神经网络输出值的矩阵计算方法。\n",
    "\n",
    "如果写成一个公式：\n",
    "$$\n",
    "\\vec{y} = f(W4 \\cdot f(W3 \\cdot f(W2 \\cdot f(W1 \\cdot \\vec{x}))))\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "神经网络正向计算的过程比较简单，就是一层一层不断做运算就可以了，动态的演示如下图所示：\n",
    "![](images/nn-forward.gif)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. 神经网络的训练 - 反向传播算法\n",
    "\n",
    "神经网络的每个连接上的权值如果知道，那么就可以将输入数据代入得到希望的结果。我们可以说神经网络是一个模型，那么这些权值就是**模型的参数**，也就是模型要学习的东西。然而，一个神经网络的连接方式、网络的层数、每层的节点数这些参数，则不是学习出来的，而是人为事先设置的。对于这些人为设置的参数，我们称之为**超参数(Hyper-Parameters)**。\n",
    "\n",
    "前面课程中所学的最小二乘、逻辑回归等可以直接优化损失函数来求解模型参数的更新值。在多层神经网络中，最后一层的参数可以用这样的方式求解得到；隐层节点没有输出的真值，因此无法直接构建损失函数来求解，如何化解这个难题？\n",
    "\n",
    "反向传播算法其实就是链式求导法则的应用。然而，这个如此简单且显而易见的方法，却是在Roseblatt提出感知器算法将近30年之后才被发明和普及的。对此，Bengio这样回应道：\n",
    "\n",
    "> 很多看似显而易见的想法只有在事后才变得显而易见。\n",
    "\n",
    "按照机器学习的通用套路，我们先确定神经网络的目标函数，然后用随机梯度下降优化算法去求目标函数最小值时的参数值。\n",
    "\n",
    "我们取网络所有输出层节点的误差平方和作为目标函数：\n",
    "\n",
    "![bp_loss](images/bp_loss.png)\n",
    "\n",
    "其中，$E_d$表示是样本$d$的误差, **t是样本的标签值**，**y是神经网络的输出值**。\n",
    "\n",
    "然后，使用随机梯度下降算法对目标函数进行优化：\n",
    "\n",
    "![bp_weight_update](images/bp_weight_update.png)\n",
    "\n",
    "随机梯度下降算法也就是需要求出误差$E_d$对于每个权重$w_{ji}$的偏导数（也就是梯度），怎么求呢？\n",
    "\n",
    "![nn3](images/nn3.png)\n",
    "\n",
    "观察上图，我们发现权重$w_{ji}$仅能通过影响节点$j$的输入值影响网络的其它部分，设$net_j$是节点$j$的加权输入，即\n",
    "\n",
    "![eqn_21_22](images/eqn_21_22.png)\n",
    "\n",
    "$E_d$是$net_j$的函数，而$net_j$是$w_{ji}$的函数。根据链式求导法则，可以得到：(FIXME: change i -> k)\n",
    "\n",
    "![eqn_23_25](images/eqn_23_25.png)\n",
    "\n",
    "\n",
    "上式中，$x_{ji}$是节点传递给节点$j$的输入值，也就是节点$i$的输出值。\n",
    "\n",
    "对于的$\\frac{\\partial E_d}{\\partial net_j}$推导，需要区分输出层和隐藏层两种情况。\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.1 输出层权值训练\n",
    "\n",
    "![nn3](images/nn3.png)\n",
    "\n",
    "对于输出层来说，$net_j$仅能通过节点$j$的输出值$y_j$来影响网络其它部分，也就是说$E_d$是$y_j$的函数，而$y_j$是$net_j$的函数，其中$y_j = sigmod(net_j)$。所以我们可以再次使用链式求导法则：\n",
    "\n",
    "![eqn_26](images/eqn_26.png)\n",
    "\n",
    "考虑上式第一项:\n",
    "\n",
    "![eqn_27_29](images/eqn_27_29.png)\n",
    "\n",
    "\n",
    "考虑上式第二项：\n",
    "\n",
    "![eqn_30_31](images/eqn_30_31.png)\n",
    "\n",
    "将第一项和第二项带入，得到：\n",
    "\n",
    "![eqn_ed_net_j.png](images/eqn_ed_net_j.png)\n",
    "\n",
    "如果令$\\delta_j = - \\frac{\\partial E_d}{\\partial net_j}$，也就是一个节点的误差项$\\delta$是网络误差对这个节点输入的偏导数的相反数。带入上式，得到：\n",
    "\n",
    "![eqn_delta_j.png](images/eqn_delta_j.png)\n",
    "\n",
    "将上述推导带入随机梯度下降公式，得到：\n",
    "\n",
    "![eqn_32_34.png](images/eqn_32_34.png)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.2 隐藏层权值训练\n",
    "\n",
    "现在我们要推导出隐藏层的$\\frac{\\partial E_d}{\\partial net_j}$。\n",
    "\n",
    "![nn3](images/nn3.png)\n",
    "\n",
    "首先，我们需要定义节点$j$的所有直接下游节点的集合$Downstream(j)$。例如，对于节点4来说，它的直接下游节点是节点8、节点9。可以看到$net_j$只能通过影响$Downstream(j)$再影响$E_d$。设$net_k$是节点$j$的下游节点的输入，则$E_d$是$net_k$的函数，而$net_k$是$net_j$的函数。因为$net_k$有多个，我们应用全导数公式，可以做出如下推导：\n",
    "\n",
    "![eqn_35_40](images/eqn_35_40.png)\n",
    "\n",
    "因为$\\delta_j = - \\frac{\\partial E_d}{\\partial net_j}$，带入上式得到：\n",
    "\n",
    "![eqn_delta_hidden.png](images/eqn_delta_hidden.png)\n",
    "\n",
    "\n",
    "至此，我们已经推导出了反向传播算法。需要注意的是，我们刚刚推导出的训练规则是根据激活函数是sigmoid函数、平方和误差、全连接网络、随机梯度下降优化算法。如果激活函数不同、误差计算方式不同、网络连接结构不同、优化算法不同，则具体的训练规则也会不一样。但是无论怎样，训练规则的推导方式都是一样的，应用链式求导法则进行推导即可。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  5.3 具体解释\n",
    "\n",
    "我们假设每个训练样本为$(\\vec{x}, \\vec{t})$，其中向量$\\vec{x}$是训练样本的特征，而$\\vec{t}$是样本的目标值。\n",
    "\n",
    "![nn3](images/nn3.png)\n",
    "\n",
    "首先，我们根据上一节介绍的算法，用样本的特征$\\vec{x}$，计算出神经网络中每个隐藏层节点的输出$a_i$，以及输出层每个节点的输出$y_i$。\n",
    "\n",
    "然后，我们按照下面的方法计算出每个节点的误差项$\\delta_i$：\n",
    "\n",
    "* **对于输出层节点$i$**\n",
    "\n",
    "![formular_3.png](images/formular_3.png)\n",
    "\n",
    "其中，$\\delta_i$是节点$i$的误差项，$y_i$是节点$i$的输出值，$t_i$是样本对应于节点$i$的目标值。举个例子，根据上图，对于输出层节点8来说，它的输出值是$y_1$，而样本的目标值是$t_1$，带入上面的公式得到节点8的误差项应该是：\n",
    "\n",
    "![forumlar_delta8.png](images/forumlar_delta8.png)\n",
    "\n",
    "* **对于隐藏层节点**\n",
    "\n",
    "![formular_4.png](images/formular_4.png)\n",
    "\n",
    "其中，$a_i$是节点$i$的输出值，$w_{ki}$是节点$i$到它的下一层节点$k$的连接的权重，$\\delta_k$是节点$i$的下一层节点$k$的误差项。例如，对于隐藏层节点4来说，计算方法如下：\n",
    "\n",
    "![forumlar_delta4.png](images/forumlar_delta4.png)\n",
    "\n",
    "\n",
    "\n",
    "最后，更新每个连接上的权值：\n",
    "\n",
    "![formular_5.png](images/formular_5.png)\n",
    "\n",
    "其中，$w_{ji}$是节点$i$到节点$j$的权重，$\\eta$是一个成为学习速率的常数，$\\delta_j$是节点$j$的误差项，$x_{ji}$是节点$i$传递给节点$j$的输入。例如，权重$w_{84}$的更新方法如下：\n",
    "\n",
    "![eqn_w84_update.png](images/eqn_w84_update.png)\n",
    "\n",
    "类似的，权重$w_{41}$的更新方法如下：\n",
    "\n",
    "![eqn_w41_update.png](images/eqn_w41_update.png)\n",
    "\n",
    "\n",
    "偏置项的输入值永远为1。例如，节点4的偏置项$w_{4b}$应该按照下面的方法计算：\n",
    "\n",
    "![eqn_w4b_update.png](images/eqn_w4b_update.png)\n",
    "\n",
    "我们已经介绍了神经网络每个节点误差项的计算和权重更新方法。显然，计算一个节点的误差项，需要先计算每个与其相连的下一层节点的误差项。这就要求误差项的计算顺序必须是从输出层开始，然后反向依次计算每个隐藏层的误差项，直到与输入层相连的那个隐藏层。这就是反向传播算法的名字的含义。当所有节点的误差项计算完毕后，我们就可以根据式5来更新所有的权重。\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. 为什么要使用激活函数\n",
    "激活函数在神经网络中非常重要，使用激活函数也是非常必要的，前面我们从人脑神经元的角度理解了激活函数，因为神经元需要通过激活才能往后传播，所以神经网络中需要激活函数，下面我们从数学的角度理解一下激活函数的必要性。\n",
    "\n",
    "比如一个两层的神经网络，使用 A 表示激活函数，那么\n",
    "\n",
    "$$\n",
    "y = w_2 A(w_1 x)\n",
    "$$\n",
    "\n",
    "如果我们不使用激活函数，那么神经网络的结果就是\n",
    "\n",
    "$$\n",
    "y = w_2 (w_1 x) = (w_2 w_1) x = \\bar{w} x\n",
    "$$\n",
    "\n",
    "可以看到，我们将两层神经网络的参数合在一起，用 $\\bar{w}$ 来表示，两层的神经网络其实就变成了一层神经网络，只不过参数变成了新的 $\\bar{w}$，所以如果不使用激活函数，那么不管多少层的神经网络，$y = w_n \\cdots w_2 w_1 x = \\bar{w} x$，就都变成了单层神经网络，所以在每一层我们都必须使用激活函数。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "最后我们看看激活函数对神经网络的影响\n",
    "\n",
    "![](images/nn-activation-function.gif)\n",
    "\n",
    "可以看到使用了激活函数之后，神经网络可以通过改变权重实现任意形状，越是复杂的神经网络能拟合的形状越复杂，这就是著名的神经网络万有逼近定理。神经网络使用的激活函数都是非线性的，每个激活函数都输入一个值，然后做一种特定的数学运算得到一个结果。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.1 sigmoid 激活函数\n",
    "\n",
    "$$\\sigma(x) = \\frac{1}{1 + e^{-x}}$$\n",
    "\n",
    "![](images/act-sigmoid.jpg)\n",
    "\n",
    "### 6.2 tanh 激活函数\n",
    "\n",
    "$$tanh(x) = 2 \\sigma(2x) - 1$$\n",
    "\n",
    "![](images/act-tanh.jpg)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.3 ReLU 激活函数\n",
    "\n",
    "$$ReLU(x) = max(0, x)$$\n",
    "\n",
    "![](images/act-relu.jpg)\n",
    "\n",
    "当输入 $x<0$ 时，输出为 $0$，当 $x> 0$ 时，输出为 $x$。该激活函数使网络更快速地收敛。它不会饱和，即它可以对抗梯度消失问题，至少在正区域（$x> 0$ 时）可以这样，因此神经元至少在一半区域中不会把所有零进行反向传播。由于使用了简单的阈值化（thresholding），ReLU 计算效率很高。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在网络中，不同的输入可能包含着大小不同关键特征，使用大小可变的数据结构去做容器，则更加灵活。假如神经元激活具有稀疏性，那么不同激活路径上：不同数量（选择性不激活）、不同功能（分布式激活）。两种可优化的结构生成的激活路径，可以更好地从有效的数据的维度上，学习到相对稀疏的特征，起到自动化解离效果。\n",
    "\n",
    "![](images/nn-sparse.png)\n",
    "\n",
    "稀疏特征并不需要网络具有很强的处理线性不可分机制，因此在深度学习模型中，使用简单、速度快的线性激活函数可能更为合适。如图，一旦神经元与神经元之间改为线性激活，网络的非线性部分仅仅来自于神经元部分选择性激活。\n",
    "\n",
    "\n",
    "更倾向于使用线性神经激活函数的另外一个原因是，减轻梯度法训练深度网络时的Vanishing Gradient Problem。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "看过BP推导的人都知道，误差从输出层反向传播算梯度时，在各层都要乘当前层的输入神经元值，激活函数的一阶导数。\n",
    "$$\n",
    "grad = error ⋅ sigmoid'(x) ⋅ x\n",
    "$$\n",
    "\n",
    "使用双端饱和(即值域被限制)Sigmoid系函数会有两个问题：\n",
    "\n",
    "1. sigmoid'(x) ∈ (0,1)  导数缩放\n",
    "2. x∈(0,1)或x∈(-1,1)  饱和值缩放\n",
    "\n",
    "这样，经过每一层时，Error都是成倍的衰减，一旦进行递推式的多层的反向传播，梯度就会不停的衰减，消失，使得网络学习变慢。而校正激活函数的梯度是1，且只有一端饱和，梯度很好的在反向传播中流动，训练速度得到了很大的提高。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. 算法与处理步骤\n",
    "\n",
    "```\n",
    "# 每次训练\n",
    "for k in range(epoch)\n",
    "    # 正向计算\n",
    "    for j in range(NN_depth):\n",
    "        # 式2 ( a = xxx)\n",
    "        X_j = f( W_{j, j-1} X_{j-1})\n",
    "\n",
    "    ＃ 反向误差计算\n",
    "    for j in range(NN_depth, 0, -1):\n",
    "        # 式3， 式4\n",
    "        delta = y_i(1-y_i)(t_i-y_i)\n",
    "        or \n",
    "        delta = a_i(1-a_i) \\sum w_ki delta_k\n",
    "\n",
    "        # 式5\n",
    "        w_ji = w_j + epsilon delta_j x_ji\n",
    "```\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8. 示例程序"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "from sklearn import datasets, linear_model\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# generate sample data\n",
    "np.random.seed(0)\n",
    "X, y = datasets.make_moons(200, noise=0.20)\n",
    "\n",
    "# generate nn output target\n",
    "t = np.zeros((X.shape[0], 2))\n",
    "t[np.where(y==0), 0] = 1\n",
    "t[np.where(y==1), 1] = 1\n",
    "\n",
    "# plot data\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y, cmap=plt.cm.Spectral)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# generate the NN model\n",
    "class NN_Model:\n",
    "    epsilon = 0.01               # learning rate\n",
    "    n_epoch = 1000               # iterative number\n",
    "    \n",
    "nn = NN_Model()\n",
    "nn.n_input_dim = X.shape[1]      # input size\n",
    "nn.n_output_dim = 2              # output node size\n",
    "nn.n_hide_dim = 8                # hidden node size\n",
    "\n",
    "nn.X = X\n",
    "nn.y = y    \n",
    "\n",
    "# initial weight array\n",
    "nn.W1 = np.random.randn(nn.n_input_dim, nn.n_hide_dim) / np.sqrt(nn.n_input_dim)\n",
    "nn.b1 = np.zeros((1, nn.n_hide_dim))\n",
    "nn.W2 = np.random.randn(nn.n_hide_dim, nn.n_output_dim) / np.sqrt(nn.n_hide_dim)\n",
    "nn.b2 = np.zeros((1, nn.n_output_dim))\n",
    "\n",
    "# define sigmod & its derivate function\n",
    "def sigmod(X):\n",
    "    return 1.0/(1+np.exp(-X))\n",
    "\n",
    "# network forward calculation\n",
    "def forward(n, X):\n",
    "    n.z1 = sigmod(X.dot(n.W1) + n.b1)\n",
    "    n.z2 = sigmod(n.z1.dot(n.W2) + n.b2)\n",
    "    return n\n",
    "\n",
    "\n",
    "# use random weight to perdict\n",
    "forward(nn, X)\n",
    "y_pred = np.argmax(nn.z2, axis=1)\n",
    "\n",
    "# plot data\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y_pred, cmap=plt.cm.Spectral)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [   0] L = 109.763265, acc = 0.500000\n",
      "epoch [   1] L = 103.996033, acc = 0.500000\n",
      "epoch [   2] L = 100.061412, acc = 0.500000\n",
      "epoch [   3] L = 97.202024, acc = 0.615000\n",
      "epoch [   4] L = 94.891877, acc = 0.815000\n",
      "epoch [   5] L = 92.856687, acc = 0.805000\n",
      "epoch [   6] L = 90.969326, acc = 0.800000\n",
      "epoch [   7] L = 89.173360, acc = 0.795000\n",
      "epoch [   8] L = 87.443892, acc = 0.795000\n",
      "epoch [   9] L = 85.769759, acc = 0.790000\n",
      "epoch [  10] L = 84.145805, acc = 0.785000\n",
      "epoch [  11] L = 82.569580, acc = 0.795000\n",
      "epoch [  12] L = 81.039921, acc = 0.800000\n",
      "epoch [  13] L = 79.556306, acc = 0.800000\n",
      "epoch [  14] L = 78.118541, acc = 0.800000\n",
      "epoch [  15] L = 76.726582, acc = 0.800000\n",
      "epoch [  16] L = 75.380423, acc = 0.795000\n",
      "epoch [  17] L = 74.080019, acc = 0.790000\n",
      "epoch [  18] L = 72.825231, acc = 0.800000\n",
      "epoch [  19] L = 71.615794, acc = 0.790000\n",
      "epoch [  20] L = 70.451292, acc = 0.795000\n",
      "epoch [  21] L = 69.331151, acc = 0.795000\n",
      "epoch [  22] L = 68.254644, acc = 0.795000\n",
      "epoch [  23] L = 67.220892, acc = 0.795000\n",
      "epoch [  24] L = 66.228884, acc = 0.795000\n",
      "epoch [  25] L = 65.277491, acc = 0.795000\n",
      "epoch [  26] L = 64.365486, acc = 0.800000\n",
      "epoch [  27] L = 63.491567, acc = 0.800000\n",
      "epoch [  28] L = 62.654371, acc = 0.800000\n",
      "epoch [  29] L = 61.852496, acc = 0.805000\n",
      "epoch [  30] L = 61.084520, acc = 0.805000\n",
      "epoch [  31] L = 60.349012, acc = 0.805000\n",
      "epoch [  32] L = 59.644549, acc = 0.805000\n",
      "epoch [  33] L = 58.969725, acc = 0.805000\n",
      "epoch [  34] L = 58.323162, acc = 0.805000\n",
      "epoch [  35] L = 57.703518, acc = 0.810000\n",
      "epoch [  36] L = 57.109490, acc = 0.820000\n",
      "epoch [  37] L = 56.539824, acc = 0.820000\n",
      "epoch [  38] L = 55.993312, acc = 0.820000\n",
      "epoch [  39] L = 55.468799, acc = 0.820000\n",
      "epoch [  40] L = 54.965181, acc = 0.820000\n",
      "epoch [  41] L = 54.481407, acc = 0.825000\n",
      "epoch [  42] L = 54.016480, acc = 0.825000\n",
      "epoch [  43] L = 53.569452, acc = 0.825000\n",
      "epoch [  44] L = 53.139426, acc = 0.825000\n",
      "epoch [  45] L = 52.725557, acc = 0.825000\n",
      "epoch [  46] L = 52.327042, acc = 0.825000\n",
      "epoch [  47] L = 51.943128, acc = 0.825000\n",
      "epoch [  48] L = 51.573103, acc = 0.825000\n",
      "epoch [  49] L = 51.216296, acc = 0.825000\n",
      "epoch [  50] L = 50.872075, acc = 0.825000\n",
      "epoch [  51] L = 50.539847, acc = 0.825000\n",
      "epoch [  52] L = 50.219052, acc = 0.825000\n",
      "epoch [  53] L = 49.909163, acc = 0.830000\n",
      "epoch [  54] L = 49.609684, acc = 0.830000\n",
      "epoch [  55] L = 49.320150, acc = 0.830000\n",
      "epoch [  56] L = 49.040121, acc = 0.830000\n",
      "epoch [  57] L = 48.769183, acc = 0.830000\n",
      "epoch [  58] L = 48.506946, acc = 0.830000\n",
      "epoch [  59] L = 48.253043, acc = 0.830000\n",
      "epoch [  60] L = 48.007127, acc = 0.830000\n",
      "epoch [  61] L = 47.768872, acc = 0.830000\n",
      "epoch [  62] L = 47.537968, acc = 0.830000\n",
      "epoch [  63] L = 47.314124, acc = 0.830000\n",
      "epoch [  64] L = 47.097064, acc = 0.830000\n",
      "epoch [  65] L = 46.886526, acc = 0.830000\n",
      "epoch [  66] L = 46.682264, acc = 0.830000\n",
      "epoch [  67] L = 46.484042, acc = 0.830000\n",
      "epoch [  68] L = 46.291638, acc = 0.830000\n",
      "epoch [  69] L = 46.104842, acc = 0.830000\n",
      "epoch [  70] L = 45.923452, acc = 0.835000\n",
      "epoch [  71] L = 45.747276, acc = 0.835000\n",
      "epoch [  72] L = 45.576134, acc = 0.835000\n",
      "epoch [  73] L = 45.409851, acc = 0.835000\n",
      "epoch [  74] L = 45.248263, acc = 0.835000\n",
      "epoch [  75] L = 45.091210, acc = 0.835000\n",
      "epoch [  76] L = 44.938543, acc = 0.835000\n",
      "epoch [  77] L = 44.790116, acc = 0.835000\n",
      "epoch [  78] L = 44.645792, acc = 0.835000\n",
      "epoch [  79] L = 44.505437, acc = 0.835000\n",
      "epoch [  80] L = 44.368925, acc = 0.835000\n",
      "epoch [  81] L = 44.236133, acc = 0.835000\n",
      "epoch [  82] L = 44.106944, acc = 0.835000\n",
      "epoch [  83] L = 43.981245, acc = 0.835000\n",
      "epoch [  84] L = 43.858928, acc = 0.835000\n",
      "epoch [  85] L = 43.739889, acc = 0.835000\n",
      "epoch [  86] L = 43.624027, acc = 0.835000\n",
      "epoch [  87] L = 43.511245, acc = 0.835000\n",
      "epoch [  88] L = 43.401450, acc = 0.835000\n",
      "epoch [  89] L = 43.294551, acc = 0.835000\n",
      "epoch [  90] L = 43.190461, acc = 0.835000\n",
      "epoch [  91] L = 43.089097, acc = 0.835000\n",
      "epoch [  92] L = 42.990376, acc = 0.835000\n",
      "epoch [  93] L = 42.894222, acc = 0.835000\n",
      "epoch [  94] L = 42.800557, acc = 0.835000\n",
      "epoch [  95] L = 42.709308, acc = 0.835000\n",
      "epoch [  96] L = 42.620404, acc = 0.830000\n",
      "epoch [  97] L = 42.533777, acc = 0.835000\n",
      "epoch [  98] L = 42.449360, acc = 0.835000\n",
      "epoch [  99] L = 42.367088, acc = 0.835000\n",
      "epoch [ 100] L = 42.286900, acc = 0.835000\n",
      "epoch [ 101] L = 42.208734, acc = 0.835000\n",
      "epoch [ 102] L = 42.132533, acc = 0.835000\n",
      "epoch [ 103] L = 42.058239, acc = 0.835000\n",
      "epoch [ 104] L = 41.985798, acc = 0.835000\n",
      "epoch [ 105] L = 41.915156, acc = 0.835000\n",
      "epoch [ 106] L = 41.846262, acc = 0.835000\n",
      "epoch [ 107] L = 41.779066, acc = 0.835000\n",
      "epoch [ 108] L = 41.713520, acc = 0.835000\n",
      "epoch [ 109] L = 41.649576, acc = 0.835000\n",
      "epoch [ 110] L = 41.587189, acc = 0.835000\n",
      "epoch [ 111] L = 41.526315, acc = 0.835000\n",
      "epoch [ 112] L = 41.466911, acc = 0.835000\n",
      "epoch [ 113] L = 41.408936, acc = 0.835000\n",
      "epoch [ 114] L = 41.352349, acc = 0.835000\n",
      "epoch [ 115] L = 41.297112, acc = 0.835000\n",
      "epoch [ 116] L = 41.243187, acc = 0.835000\n",
      "epoch [ 117] L = 41.190536, acc = 0.835000\n",
      "epoch [ 118] L = 41.139125, acc = 0.835000\n",
      "epoch [ 119] L = 41.088920, acc = 0.835000\n",
      "epoch [ 120] L = 41.039886, acc = 0.835000\n",
      "epoch [ 121] L = 40.991992, acc = 0.835000\n",
      "epoch [ 122] L = 40.945205, acc = 0.835000\n",
      "epoch [ 123] L = 40.899496, acc = 0.835000\n",
      "epoch [ 124] L = 40.854835, acc = 0.835000\n",
      "epoch [ 125] L = 40.811194, acc = 0.835000\n",
      "epoch [ 126] L = 40.768544, acc = 0.835000\n",
      "epoch [ 127] L = 40.726859, acc = 0.835000\n",
      "epoch [ 128] L = 40.686113, acc = 0.835000\n",
      "epoch [ 129] L = 40.646280, acc = 0.835000\n",
      "epoch [ 130] L = 40.607336, acc = 0.835000\n",
      "epoch [ 131] L = 40.569257, acc = 0.840000\n",
      "epoch [ 132] L = 40.532020, acc = 0.840000\n",
      "epoch [ 133] L = 40.495602, acc = 0.840000\n",
      "epoch [ 134] L = 40.459982, acc = 0.840000\n",
      "epoch [ 135] L = 40.425138, acc = 0.840000\n",
      "epoch [ 136] L = 40.391051, acc = 0.840000\n",
      "epoch [ 137] L = 40.357701, acc = 0.840000\n",
      "epoch [ 138] L = 40.325067, acc = 0.840000\n",
      "epoch [ 139] L = 40.293132, acc = 0.840000\n",
      "epoch [ 140] L = 40.261877, acc = 0.840000\n",
      "epoch [ 141] L = 40.231285, acc = 0.845000\n",
      "epoch [ 142] L = 40.201338, acc = 0.845000\n",
      "epoch [ 143] L = 40.172021, acc = 0.845000\n",
      "epoch [ 144] L = 40.143316, acc = 0.845000\n",
      "epoch [ 145] L = 40.115210, acc = 0.845000\n",
      "epoch [ 146] L = 40.087685, acc = 0.845000\n",
      "epoch [ 147] L = 40.060728, acc = 0.845000\n",
      "epoch [ 148] L = 40.034324, acc = 0.850000\n",
      "epoch [ 149] L = 40.008459, acc = 0.850000\n",
      "epoch [ 150] L = 39.983120, acc = 0.850000\n",
      "epoch [ 151] L = 39.958295, acc = 0.850000\n",
      "epoch [ 152] L = 39.933969, acc = 0.850000\n",
      "epoch [ 153] L = 39.910131, acc = 0.850000\n",
      "epoch [ 154] L = 39.886769, acc = 0.850000\n",
      "epoch [ 155] L = 39.863871, acc = 0.850000\n",
      "epoch [ 156] L = 39.841426, acc = 0.855000\n",
      "epoch [ 157] L = 39.819423, acc = 0.855000\n",
      "epoch [ 158] L = 39.797851, acc = 0.855000\n",
      "epoch [ 159] L = 39.776699, acc = 0.855000\n",
      "epoch [ 160] L = 39.755959, acc = 0.855000\n",
      "epoch [ 161] L = 39.735619, acc = 0.855000\n",
      "epoch [ 162] L = 39.715671, acc = 0.855000\n",
      "epoch [ 163] L = 39.696104, acc = 0.855000\n",
      "epoch [ 164] L = 39.676911, acc = 0.855000\n",
      "epoch [ 165] L = 39.658082, acc = 0.855000\n",
      "epoch [ 166] L = 39.639609, acc = 0.855000\n",
      "epoch [ 167] L = 39.621483, acc = 0.855000\n",
      "epoch [ 168] L = 39.603696, acc = 0.855000\n",
      "epoch [ 169] L = 39.586241, acc = 0.855000\n",
      "epoch [ 170] L = 39.569110, acc = 0.855000\n",
      "epoch [ 171] L = 39.552296, acc = 0.855000\n",
      "epoch [ 172] L = 39.535790, acc = 0.855000\n",
      "epoch [ 173] L = 39.519587, acc = 0.855000\n",
      "epoch [ 174] L = 39.503679, acc = 0.855000\n",
      "epoch [ 175] L = 39.488060, acc = 0.855000\n",
      "epoch [ 176] L = 39.472722, acc = 0.855000\n",
      "epoch [ 177] L = 39.457661, acc = 0.855000\n",
      "epoch [ 178] L = 39.442869, acc = 0.855000\n",
      "epoch [ 179] L = 39.428341, acc = 0.855000\n",
      "epoch [ 180] L = 39.414071, acc = 0.855000\n",
      "epoch [ 181] L = 39.400052, acc = 0.855000\n",
      "epoch [ 182] L = 39.386281, acc = 0.855000\n",
      "epoch [ 183] L = 39.372750, acc = 0.855000\n",
      "epoch [ 184] L = 39.359456, acc = 0.855000\n",
      "epoch [ 185] L = 39.346392, acc = 0.855000\n",
      "epoch [ 186] L = 39.333554, acc = 0.855000\n",
      "epoch [ 187] L = 39.320937, acc = 0.855000\n",
      "epoch [ 188] L = 39.308536, acc = 0.855000\n",
      "epoch [ 189] L = 39.296346, acc = 0.855000\n",
      "epoch [ 190] L = 39.284364, acc = 0.855000\n",
      "epoch [ 191] L = 39.272584, acc = 0.855000\n",
      "epoch [ 192] L = 39.261002, acc = 0.855000\n",
      "epoch [ 193] L = 39.249614, acc = 0.855000\n",
      "epoch [ 194] L = 39.238416, acc = 0.855000\n",
      "epoch [ 195] L = 39.227405, acc = 0.855000\n",
      "epoch [ 196] L = 39.216575, acc = 0.855000\n",
      "epoch [ 197] L = 39.205924, acc = 0.855000\n",
      "epoch [ 198] L = 39.195447, acc = 0.855000\n",
      "epoch [ 199] L = 39.185142, acc = 0.855000\n",
      "epoch [ 200] L = 39.175003, acc = 0.855000\n",
      "epoch [ 201] L = 39.165029, acc = 0.855000\n",
      "epoch [ 202] L = 39.155216, acc = 0.855000\n",
      "epoch [ 203] L = 39.145560, acc = 0.855000\n",
      "epoch [ 204] L = 39.136058, acc = 0.855000\n",
      "epoch [ 205] L = 39.126707, acc = 0.855000\n",
      "epoch [ 206] L = 39.117504, acc = 0.855000\n",
      "epoch [ 207] L = 39.108446, acc = 0.855000\n",
      "epoch [ 208] L = 39.099530, acc = 0.855000\n",
      "epoch [ 209] L = 39.090753, acc = 0.855000\n",
      "epoch [ 210] L = 39.082113, acc = 0.855000\n",
      "epoch [ 211] L = 39.073606, acc = 0.855000\n",
      "epoch [ 212] L = 39.065230, acc = 0.855000\n",
      "epoch [ 213] L = 39.056983, acc = 0.855000\n",
      "epoch [ 214] L = 39.048862, acc = 0.855000\n",
      "epoch [ 215] L = 39.040864, acc = 0.855000\n",
      "epoch [ 216] L = 39.032987, acc = 0.855000\n",
      "epoch [ 217] L = 39.025229, acc = 0.855000\n",
      "epoch [ 218] L = 39.017587, acc = 0.855000\n",
      "epoch [ 219] L = 39.010059, acc = 0.855000\n",
      "epoch [ 220] L = 39.002643, acc = 0.855000\n",
      "epoch [ 221] L = 38.995337, acc = 0.855000\n",
      "epoch [ 222] L = 38.988138, acc = 0.855000\n",
      "epoch [ 223] L = 38.981045, acc = 0.855000\n",
      "epoch [ 224] L = 38.974055, acc = 0.855000\n",
      "epoch [ 225] L = 38.967166, acc = 0.855000\n",
      "epoch [ 226] L = 38.960377, acc = 0.855000\n",
      "epoch [ 227] L = 38.953686, acc = 0.855000\n",
      "epoch [ 228] L = 38.947090, acc = 0.855000\n",
      "epoch [ 229] L = 38.940588, acc = 0.855000\n",
      "epoch [ 230] L = 38.934178, acc = 0.855000\n",
      "epoch [ 231] L = 38.927859, acc = 0.855000\n",
      "epoch [ 232] L = 38.921628, acc = 0.855000\n",
      "epoch [ 233] L = 38.915484, acc = 0.855000\n",
      "epoch [ 234] L = 38.909426, acc = 0.855000\n",
      "epoch [ 235] L = 38.903452, acc = 0.855000\n",
      "epoch [ 236] L = 38.897559, acc = 0.855000\n",
      "epoch [ 237] L = 38.891747, acc = 0.855000\n",
      "epoch [ 238] L = 38.886015, acc = 0.855000\n",
      "epoch [ 239] L = 38.880360, acc = 0.855000\n",
      "epoch [ 240] L = 38.874781, acc = 0.855000\n",
      "epoch [ 241] L = 38.869278, acc = 0.855000\n",
      "epoch [ 242] L = 38.863847, acc = 0.855000\n",
      "epoch [ 243] L = 38.858489, acc = 0.855000\n",
      "epoch [ 244] L = 38.853201, acc = 0.855000\n",
      "epoch [ 245] L = 38.847983, acc = 0.855000\n",
      "epoch [ 246] L = 38.842833, acc = 0.855000\n",
      "epoch [ 247] L = 38.837750, acc = 0.855000\n",
      "epoch [ 248] L = 38.832733, acc = 0.855000\n",
      "epoch [ 249] L = 38.827780, acc = 0.855000\n",
      "epoch [ 250] L = 38.822891, acc = 0.855000\n",
      "epoch [ 251] L = 38.818063, acc = 0.855000\n",
      "epoch [ 252] L = 38.813297, acc = 0.855000\n",
      "epoch [ 253] L = 38.808591, acc = 0.855000\n",
      "epoch [ 254] L = 38.803943, acc = 0.855000\n",
      "epoch [ 255] L = 38.799354, acc = 0.855000\n",
      "epoch [ 256] L = 38.794820, acc = 0.855000\n",
      "epoch [ 257] L = 38.790343, acc = 0.855000\n",
      "epoch [ 258] L = 38.785920, acc = 0.855000\n",
      "epoch [ 259] L = 38.781552, acc = 0.855000\n",
      "epoch [ 260] L = 38.777235, acc = 0.855000\n",
      "epoch [ 261] L = 38.772971, acc = 0.855000\n",
      "epoch [ 262] L = 38.768757, acc = 0.855000\n",
      "epoch [ 263] L = 38.764594, acc = 0.855000\n",
      "epoch [ 264] L = 38.760479, acc = 0.855000\n",
      "epoch [ 265] L = 38.756413, acc = 0.855000\n",
      "epoch [ 266] L = 38.752394, acc = 0.855000\n",
      "epoch [ 267] L = 38.748421, acc = 0.855000\n",
      "epoch [ 268] L = 38.744494, acc = 0.855000\n",
      "epoch [ 269] L = 38.740612, acc = 0.855000\n",
      "epoch [ 270] L = 38.736774, acc = 0.855000\n",
      "epoch [ 271] L = 38.732979, acc = 0.855000\n",
      "epoch [ 272] L = 38.729227, acc = 0.855000\n",
      "epoch [ 273] L = 38.725516, acc = 0.855000\n",
      "epoch [ 274] L = 38.721846, acc = 0.855000\n",
      "epoch [ 275] L = 38.718217, acc = 0.855000\n",
      "epoch [ 276] L = 38.714626, acc = 0.855000\n",
      "epoch [ 277] L = 38.711075, acc = 0.855000\n",
      "epoch [ 278] L = 38.707562, acc = 0.855000\n",
      "epoch [ 279] L = 38.704087, acc = 0.855000\n",
      "epoch [ 280] L = 38.700648, acc = 0.855000\n",
      "epoch [ 281] L = 38.697245, acc = 0.855000\n",
      "epoch [ 282] L = 38.693878, acc = 0.855000\n",
      "epoch [ 283] L = 38.690545, acc = 0.855000\n",
      "epoch [ 284] L = 38.687247, acc = 0.855000\n",
      "epoch [ 285] L = 38.683983, acc = 0.855000\n",
      "epoch [ 286] L = 38.680751, acc = 0.855000\n",
      "epoch [ 287] L = 38.677552, acc = 0.855000\n",
      "epoch [ 288] L = 38.674385, acc = 0.855000\n",
      "epoch [ 289] L = 38.671249, acc = 0.855000\n",
      "epoch [ 290] L = 38.668144, acc = 0.855000\n",
      "epoch [ 291] L = 38.665069, acc = 0.855000\n",
      "epoch [ 292] L = 38.662024, acc = 0.855000\n",
      "epoch [ 293] L = 38.659008, acc = 0.855000\n",
      "epoch [ 294] L = 38.656020, acc = 0.855000\n",
      "epoch [ 295] L = 38.653061, acc = 0.855000\n",
      "epoch [ 296] L = 38.650129, acc = 0.855000\n",
      "epoch [ 297] L = 38.647224, acc = 0.855000\n",
      "epoch [ 298] L = 38.644346, acc = 0.855000\n",
      "epoch [ 299] L = 38.641494, acc = 0.855000\n",
      "epoch [ 300] L = 38.638668, acc = 0.855000\n",
      "epoch [ 301] L = 38.635867, acc = 0.855000\n",
      "epoch [ 302] L = 38.633091, acc = 0.855000\n",
      "epoch [ 303] L = 38.630339, acc = 0.855000\n",
      "epoch [ 304] L = 38.627611, acc = 0.855000\n",
      "epoch [ 305] L = 38.624906, acc = 0.855000\n",
      "epoch [ 306] L = 38.622225, acc = 0.855000\n",
      "epoch [ 307] L = 38.619566, acc = 0.855000\n",
      "epoch [ 308] L = 38.616929, acc = 0.855000\n",
      "epoch [ 309] L = 38.614314, acc = 0.855000\n",
      "epoch [ 310] L = 38.611720, acc = 0.855000\n",
      "epoch [ 311] L = 38.609148, acc = 0.855000\n",
      "epoch [ 312] L = 38.606596, acc = 0.855000\n",
      "epoch [ 313] L = 38.604064, acc = 0.855000\n",
      "epoch [ 314] L = 38.601552, acc = 0.855000\n",
      "epoch [ 315] L = 38.599060, acc = 0.855000\n",
      "epoch [ 316] L = 38.596587, acc = 0.855000\n",
      "epoch [ 317] L = 38.594133, acc = 0.855000\n",
      "epoch [ 318] L = 38.591697, acc = 0.855000\n",
      "epoch [ 319] L = 38.589279, acc = 0.855000\n",
      "epoch [ 320] L = 38.586879, acc = 0.855000\n",
      "epoch [ 321] L = 38.584497, acc = 0.855000\n",
      "epoch [ 322] L = 38.582131, acc = 0.855000\n",
      "epoch [ 323] L = 38.579783, acc = 0.855000\n",
      "epoch [ 324] L = 38.577451, acc = 0.855000\n",
      "epoch [ 325] L = 38.575135, acc = 0.855000\n",
      "epoch [ 326] L = 38.572835, acc = 0.855000\n",
      "epoch [ 327] L = 38.570550, acc = 0.855000\n",
      "epoch [ 328] L = 38.568281, acc = 0.855000\n",
      "epoch [ 329] L = 38.566027, acc = 0.855000\n",
      "epoch [ 330] L = 38.563787, acc = 0.855000\n",
      "epoch [ 331] L = 38.561562, acc = 0.855000\n",
      "epoch [ 332] L = 38.559351, acc = 0.855000\n",
      "epoch [ 333] L = 38.557154, acc = 0.855000\n",
      "epoch [ 334] L = 38.554970, acc = 0.855000\n",
      "epoch [ 335] L = 38.552800, acc = 0.855000\n",
      "epoch [ 336] L = 38.550643, acc = 0.855000\n",
      "epoch [ 337] L = 38.548498, acc = 0.855000\n",
      "epoch [ 338] L = 38.546366, acc = 0.855000\n",
      "epoch [ 339] L = 38.544247, acc = 0.855000\n",
      "epoch [ 340] L = 38.542139, acc = 0.855000\n",
      "epoch [ 341] L = 38.540043, acc = 0.855000\n",
      "epoch [ 342] L = 38.537959, acc = 0.855000\n",
      "epoch [ 343] L = 38.535886, acc = 0.855000\n",
      "epoch [ 344] L = 38.533824, acc = 0.855000\n",
      "epoch [ 345] L = 38.531773, acc = 0.855000\n",
      "epoch [ 346] L = 38.529733, acc = 0.855000\n",
      "epoch [ 347] L = 38.527703, acc = 0.855000\n",
      "epoch [ 348] L = 38.525683, acc = 0.855000\n",
      "epoch [ 349] L = 38.523673, acc = 0.855000\n",
      "epoch [ 350] L = 38.521673, acc = 0.855000\n",
      "epoch [ 351] L = 38.519682, acc = 0.855000\n",
      "epoch [ 352] L = 38.517701, acc = 0.855000\n",
      "epoch [ 353] L = 38.515729, acc = 0.855000\n",
      "epoch [ 354] L = 38.513766, acc = 0.860000\n",
      "epoch [ 355] L = 38.511812, acc = 0.860000\n",
      "epoch [ 356] L = 38.509866, acc = 0.860000\n",
      "epoch [ 357] L = 38.507928, acc = 0.860000\n",
      "epoch [ 358] L = 38.505999, acc = 0.860000\n",
      "epoch [ 359] L = 38.504077, acc = 0.860000\n",
      "epoch [ 360] L = 38.502164, acc = 0.860000\n",
      "epoch [ 361] L = 38.500258, acc = 0.860000\n",
      "epoch [ 362] L = 38.498359, acc = 0.860000\n",
      "epoch [ 363] L = 38.496467, acc = 0.860000\n",
      "epoch [ 364] L = 38.494583, acc = 0.860000\n",
      "epoch [ 365] L = 38.492705, acc = 0.860000\n",
      "epoch [ 366] L = 38.490835, acc = 0.860000\n",
      "epoch [ 367] L = 38.488970, acc = 0.860000\n",
      "epoch [ 368] L = 38.487112, acc = 0.860000\n",
      "epoch [ 369] L = 38.485261, acc = 0.860000\n",
      "epoch [ 370] L = 38.483415, acc = 0.860000\n",
      "epoch [ 371] L = 38.481575, acc = 0.860000\n",
      "epoch [ 372] L = 38.479741, acc = 0.860000\n",
      "epoch [ 373] L = 38.477913, acc = 0.860000\n",
      "epoch [ 374] L = 38.476090, acc = 0.860000\n",
      "epoch [ 375] L = 38.474272, acc = 0.860000\n",
      "epoch [ 376] L = 38.472459, acc = 0.860000\n",
      "epoch [ 377] L = 38.470652, acc = 0.860000\n",
      "epoch [ 378] L = 38.468849, acc = 0.860000\n",
      "epoch [ 379] L = 38.467051, acc = 0.860000\n",
      "epoch [ 380] L = 38.465257, acc = 0.860000\n",
      "epoch [ 381] L = 38.463468, acc = 0.860000\n",
      "epoch [ 382] L = 38.461683, acc = 0.860000\n",
      "epoch [ 383] L = 38.459902, acc = 0.860000\n",
      "epoch [ 384] L = 38.458125, acc = 0.860000\n",
      "epoch [ 385] L = 38.456352, acc = 0.860000\n",
      "epoch [ 386] L = 38.454582, acc = 0.860000\n",
      "epoch [ 387] L = 38.452816, acc = 0.860000\n",
      "epoch [ 388] L = 38.451054, acc = 0.860000\n",
      "epoch [ 389] L = 38.449295, acc = 0.860000\n",
      "epoch [ 390] L = 38.447539, acc = 0.855000\n",
      "epoch [ 391] L = 38.445786, acc = 0.855000\n",
      "epoch [ 392] L = 38.444036, acc = 0.855000\n",
      "epoch [ 393] L = 38.442289, acc = 0.855000\n",
      "epoch [ 394] L = 38.440545, acc = 0.855000\n",
      "epoch [ 395] L = 38.438803, acc = 0.855000\n",
      "epoch [ 396] L = 38.437064, acc = 0.855000\n",
      "epoch [ 397] L = 38.435327, acc = 0.855000\n",
      "epoch [ 398] L = 38.433592, acc = 0.855000\n",
      "epoch [ 399] L = 38.431860, acc = 0.855000\n",
      "epoch [ 400] L = 38.430129, acc = 0.855000\n",
      "epoch [ 401] L = 38.428400, acc = 0.855000\n",
      "epoch [ 402] L = 38.426673, acc = 0.855000\n",
      "epoch [ 403] L = 38.424948, acc = 0.855000\n",
      "epoch [ 404] L = 38.423224, acc = 0.855000\n",
      "epoch [ 405] L = 38.421502, acc = 0.855000\n",
      "epoch [ 406] L = 38.419781, acc = 0.855000\n",
      "epoch [ 407] L = 38.418061, acc = 0.855000\n",
      "epoch [ 408] L = 38.416343, acc = 0.855000\n",
      "epoch [ 409] L = 38.414625, acc = 0.855000\n",
      "epoch [ 410] L = 38.412909, acc = 0.855000\n",
      "epoch [ 411] L = 38.411193, acc = 0.855000\n",
      "epoch [ 412] L = 38.409478, acc = 0.855000\n",
      "epoch [ 413] L = 38.407764, acc = 0.855000\n",
      "epoch [ 414] L = 38.406050, acc = 0.855000\n",
      "epoch [ 415] L = 38.404337, acc = 0.855000\n",
      "epoch [ 416] L = 38.402624, acc = 0.855000\n",
      "epoch [ 417] L = 38.400911, acc = 0.855000\n",
      "epoch [ 418] L = 38.399198, acc = 0.855000\n",
      "epoch [ 419] L = 38.397486, acc = 0.855000\n",
      "epoch [ 420] L = 38.395773, acc = 0.855000\n",
      "epoch [ 421] L = 38.394061, acc = 0.855000\n",
      "epoch [ 422] L = 38.392348, acc = 0.855000\n",
      "epoch [ 423] L = 38.390634, acc = 0.855000\n",
      "epoch [ 424] L = 38.388921, acc = 0.855000\n",
      "epoch [ 425] L = 38.387207, acc = 0.855000\n",
      "epoch [ 426] L = 38.385492, acc = 0.855000\n",
      "epoch [ 427] L = 38.383777, acc = 0.855000\n",
      "epoch [ 428] L = 38.382061, acc = 0.860000\n",
      "epoch [ 429] L = 38.380344, acc = 0.860000\n",
      "epoch [ 430] L = 38.378626, acc = 0.860000\n",
      "epoch [ 431] L = 38.376907, acc = 0.860000\n",
      "epoch [ 432] L = 38.375188, acc = 0.860000\n",
      "epoch [ 433] L = 38.373467, acc = 0.860000\n",
      "epoch [ 434] L = 38.371744, acc = 0.860000\n",
      "epoch [ 435] L = 38.370021, acc = 0.860000\n",
      "epoch [ 436] L = 38.368296, acc = 0.860000\n",
      "epoch [ 437] L = 38.366569, acc = 0.860000\n",
      "epoch [ 438] L = 38.364841, acc = 0.860000\n",
      "epoch [ 439] L = 38.363112, acc = 0.860000\n",
      "epoch [ 440] L = 38.361380, acc = 0.860000\n",
      "epoch [ 441] L = 38.359647, acc = 0.860000\n",
      "epoch [ 442] L = 38.357912, acc = 0.860000\n",
      "epoch [ 443] L = 38.356175, acc = 0.860000\n",
      "epoch [ 444] L = 38.354436, acc = 0.860000\n",
      "epoch [ 445] L = 38.352695, acc = 0.860000\n",
      "epoch [ 446] L = 38.350952, acc = 0.860000\n",
      "epoch [ 447] L = 38.349206, acc = 0.860000\n",
      "epoch [ 448] L = 38.347459, acc = 0.860000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [ 449] L = 38.345708, acc = 0.860000\n",
      "epoch [ 450] L = 38.343956, acc = 0.860000\n",
      "epoch [ 451] L = 38.342200, acc = 0.860000\n",
      "epoch [ 452] L = 38.340443, acc = 0.860000\n",
      "epoch [ 453] L = 38.338682, acc = 0.860000\n",
      "epoch [ 454] L = 38.336919, acc = 0.860000\n",
      "epoch [ 455] L = 38.335153, acc = 0.860000\n",
      "epoch [ 456] L = 38.333384, acc = 0.860000\n",
      "epoch [ 457] L = 38.331612, acc = 0.860000\n",
      "epoch [ 458] L = 38.329836, acc = 0.860000\n",
      "epoch [ 459] L = 38.328058, acc = 0.860000\n",
      "epoch [ 460] L = 38.326277, acc = 0.860000\n",
      "epoch [ 461] L = 38.324492, acc = 0.860000\n",
      "epoch [ 462] L = 38.322704, acc = 0.860000\n",
      "epoch [ 463] L = 38.320913, acc = 0.860000\n",
      "epoch [ 464] L = 38.319118, acc = 0.860000\n",
      "epoch [ 465] L = 38.317320, acc = 0.860000\n",
      "epoch [ 466] L = 38.315518, acc = 0.860000\n",
      "epoch [ 467] L = 38.313713, acc = 0.860000\n",
      "epoch [ 468] L = 38.311904, acc = 0.860000\n",
      "epoch [ 469] L = 38.310091, acc = 0.860000\n",
      "epoch [ 470] L = 38.308274, acc = 0.860000\n",
      "epoch [ 471] L = 38.306453, acc = 0.860000\n",
      "epoch [ 472] L = 38.304628, acc = 0.860000\n",
      "epoch [ 473] L = 38.302800, acc = 0.860000\n",
      "epoch [ 474] L = 38.300967, acc = 0.860000\n",
      "epoch [ 475] L = 38.299130, acc = 0.860000\n",
      "epoch [ 476] L = 38.297288, acc = 0.860000\n",
      "epoch [ 477] L = 38.295443, acc = 0.860000\n",
      "epoch [ 478] L = 38.293593, acc = 0.860000\n",
      "epoch [ 479] L = 38.291738, acc = 0.860000\n",
      "epoch [ 480] L = 38.289879, acc = 0.860000\n",
      "epoch [ 481] L = 38.288016, acc = 0.860000\n",
      "epoch [ 482] L = 38.286148, acc = 0.860000\n",
      "epoch [ 483] L = 38.284275, acc = 0.860000\n",
      "epoch [ 484] L = 38.282397, acc = 0.860000\n",
      "epoch [ 485] L = 38.280515, acc = 0.860000\n",
      "epoch [ 486] L = 38.278628, acc = 0.860000\n",
      "epoch [ 487] L = 38.276736, acc = 0.860000\n",
      "epoch [ 488] L = 38.274839, acc = 0.860000\n",
      "epoch [ 489] L = 38.272936, acc = 0.860000\n",
      "epoch [ 490] L = 38.271029, acc = 0.860000\n",
      "epoch [ 491] L = 38.269116, acc = 0.860000\n",
      "epoch [ 492] L = 38.267199, acc = 0.860000\n",
      "epoch [ 493] L = 38.265276, acc = 0.860000\n",
      "epoch [ 494] L = 38.263347, acc = 0.865000\n",
      "epoch [ 495] L = 38.261413, acc = 0.865000\n",
      "epoch [ 496] L = 38.259474, acc = 0.865000\n",
      "epoch [ 497] L = 38.257529, acc = 0.865000\n",
      "epoch [ 498] L = 38.255578, acc = 0.865000\n",
      "epoch [ 499] L = 38.253622, acc = 0.865000\n",
      "epoch [ 500] L = 38.251660, acc = 0.865000\n",
      "epoch [ 501] L = 38.249692, acc = 0.865000\n",
      "epoch [ 502] L = 38.247719, acc = 0.865000\n",
      "epoch [ 503] L = 38.245739, acc = 0.865000\n",
      "epoch [ 504] L = 38.243753, acc = 0.865000\n",
      "epoch [ 505] L = 38.241762, acc = 0.865000\n",
      "epoch [ 506] L = 38.239764, acc = 0.865000\n",
      "epoch [ 507] L = 38.237760, acc = 0.865000\n",
      "epoch [ 508] L = 38.235750, acc = 0.865000\n",
      "epoch [ 509] L = 38.233733, acc = 0.865000\n",
      "epoch [ 510] L = 38.231711, acc = 0.865000\n",
      "epoch [ 511] L = 38.229681, acc = 0.865000\n",
      "epoch [ 512] L = 38.227646, acc = 0.865000\n",
      "epoch [ 513] L = 38.225603, acc = 0.865000\n",
      "epoch [ 514] L = 38.223554, acc = 0.865000\n",
      "epoch [ 515] L = 38.221499, acc = 0.865000\n",
      "epoch [ 516] L = 38.219437, acc = 0.865000\n",
      "epoch [ 517] L = 38.217368, acc = 0.865000\n",
      "epoch [ 518] L = 38.215292, acc = 0.865000\n",
      "epoch [ 519] L = 38.213209, acc = 0.865000\n",
      "epoch [ 520] L = 38.211119, acc = 0.865000\n",
      "epoch [ 521] L = 38.209022, acc = 0.865000\n",
      "epoch [ 522] L = 38.206918, acc = 0.865000\n",
      "epoch [ 523] L = 38.204807, acc = 0.865000\n",
      "epoch [ 524] L = 38.202689, acc = 0.865000\n",
      "epoch [ 525] L = 38.200563, acc = 0.865000\n",
      "epoch [ 526] L = 38.198431, acc = 0.865000\n",
      "epoch [ 527] L = 38.196290, acc = 0.865000\n",
      "epoch [ 528] L = 38.194142, acc = 0.865000\n",
      "epoch [ 529] L = 38.191987, acc = 0.865000\n",
      "epoch [ 530] L = 38.189824, acc = 0.865000\n",
      "epoch [ 531] L = 38.187653, acc = 0.865000\n",
      "epoch [ 532] L = 38.185475, acc = 0.865000\n",
      "epoch [ 533] L = 38.183289, acc = 0.865000\n",
      "epoch [ 534] L = 38.181095, acc = 0.865000\n",
      "epoch [ 535] L = 38.178893, acc = 0.865000\n",
      "epoch [ 536] L = 38.176683, acc = 0.865000\n",
      "epoch [ 537] L = 38.174465, acc = 0.865000\n",
      "epoch [ 538] L = 38.172239, acc = 0.865000\n",
      "epoch [ 539] L = 38.170005, acc = 0.865000\n",
      "epoch [ 540] L = 38.167762, acc = 0.865000\n",
      "epoch [ 541] L = 38.165512, acc = 0.865000\n",
      "epoch [ 542] L = 38.163252, acc = 0.865000\n",
      "epoch [ 543] L = 38.160985, acc = 0.865000\n",
      "epoch [ 544] L = 38.158709, acc = 0.865000\n",
      "epoch [ 545] L = 38.156424, acc = 0.865000\n",
      "epoch [ 546] L = 38.154131, acc = 0.865000\n",
      "epoch [ 547] L = 38.151829, acc = 0.865000\n",
      "epoch [ 548] L = 38.149518, acc = 0.865000\n",
      "epoch [ 549] L = 38.147198, acc = 0.865000\n",
      "epoch [ 550] L = 38.144870, acc = 0.865000\n",
      "epoch [ 551] L = 38.142532, acc = 0.865000\n",
      "epoch [ 552] L = 38.140186, acc = 0.865000\n",
      "epoch [ 553] L = 38.137830, acc = 0.865000\n",
      "epoch [ 554] L = 38.135465, acc = 0.865000\n",
      "epoch [ 555] L = 38.133091, acc = 0.865000\n",
      "epoch [ 556] L = 38.130708, acc = 0.865000\n",
      "epoch [ 557] L = 38.128315, acc = 0.865000\n",
      "epoch [ 558] L = 38.125913, acc = 0.865000\n",
      "epoch [ 559] L = 38.123502, acc = 0.865000\n",
      "epoch [ 560] L = 38.121080, acc = 0.865000\n",
      "epoch [ 561] L = 38.118649, acc = 0.865000\n",
      "epoch [ 562] L = 38.116209, acc = 0.865000\n",
      "epoch [ 563] L = 38.113758, acc = 0.865000\n",
      "epoch [ 564] L = 38.111298, acc = 0.865000\n",
      "epoch [ 565] L = 38.108828, acc = 0.865000\n",
      "epoch [ 566] L = 38.106348, acc = 0.865000\n",
      "epoch [ 567] L = 38.103857, acc = 0.865000\n",
      "epoch [ 568] L = 38.101357, acc = 0.865000\n",
      "epoch [ 569] L = 38.098846, acc = 0.865000\n",
      "epoch [ 570] L = 38.096325, acc = 0.865000\n",
      "epoch [ 571] L = 38.093794, acc = 0.865000\n",
      "epoch [ 572] L = 38.091252, acc = 0.865000\n",
      "epoch [ 573] L = 38.088699, acc = 0.865000\n",
      "epoch [ 574] L = 38.086136, acc = 0.865000\n",
      "epoch [ 575] L = 38.083563, acc = 0.865000\n",
      "epoch [ 576] L = 38.080978, acc = 0.865000\n",
      "epoch [ 577] L = 38.078383, acc = 0.865000\n",
      "epoch [ 578] L = 38.075777, acc = 0.865000\n",
      "epoch [ 579] L = 38.073160, acc = 0.865000\n",
      "epoch [ 580] L = 38.070532, acc = 0.865000\n",
      "epoch [ 581] L = 38.067893, acc = 0.865000\n",
      "epoch [ 582] L = 38.065242, acc = 0.865000\n",
      "epoch [ 583] L = 38.062581, acc = 0.865000\n",
      "epoch [ 584] L = 38.059907, acc = 0.865000\n",
      "epoch [ 585] L = 38.057223, acc = 0.865000\n",
      "epoch [ 586] L = 38.054527, acc = 0.865000\n",
      "epoch [ 587] L = 38.051819, acc = 0.865000\n",
      "epoch [ 588] L = 38.049100, acc = 0.865000\n",
      "epoch [ 589] L = 38.046369, acc = 0.865000\n",
      "epoch [ 590] L = 38.043626, acc = 0.865000\n",
      "epoch [ 591] L = 38.040871, acc = 0.865000\n",
      "epoch [ 592] L = 38.038104, acc = 0.865000\n",
      "epoch [ 593] L = 38.035326, acc = 0.865000\n",
      "epoch [ 594] L = 38.032534, acc = 0.865000\n",
      "epoch [ 595] L = 38.029731, acc = 0.865000\n",
      "epoch [ 596] L = 38.026916, acc = 0.865000\n",
      "epoch [ 597] L = 38.024088, acc = 0.865000\n",
      "epoch [ 598] L = 38.021247, acc = 0.865000\n",
      "epoch [ 599] L = 38.018394, acc = 0.865000\n",
      "epoch [ 600] L = 38.015528, acc = 0.865000\n",
      "epoch [ 601] L = 38.012650, acc = 0.865000\n",
      "epoch [ 602] L = 38.009758, acc = 0.865000\n",
      "epoch [ 603] L = 38.006854, acc = 0.865000\n",
      "epoch [ 604] L = 38.003937, acc = 0.865000\n",
      "epoch [ 605] L = 38.001006, acc = 0.865000\n",
      "epoch [ 606] L = 37.998063, acc = 0.865000\n",
      "epoch [ 607] L = 37.995106, acc = 0.865000\n",
      "epoch [ 608] L = 37.992136, acc = 0.865000\n",
      "epoch [ 609] L = 37.989152, acc = 0.865000\n",
      "epoch [ 610] L = 37.986155, acc = 0.865000\n",
      "epoch [ 611] L = 37.983144, acc = 0.865000\n",
      "epoch [ 612] L = 37.980120, acc = 0.865000\n",
      "epoch [ 613] L = 37.977081, acc = 0.865000\n",
      "epoch [ 614] L = 37.974029, acc = 0.865000\n",
      "epoch [ 615] L = 37.970963, acc = 0.865000\n",
      "epoch [ 616] L = 37.967882, acc = 0.865000\n",
      "epoch [ 617] L = 37.964788, acc = 0.865000\n",
      "epoch [ 618] L = 37.961679, acc = 0.865000\n",
      "epoch [ 619] L = 37.958556, acc = 0.865000\n",
      "epoch [ 620] L = 37.955418, acc = 0.865000\n",
      "epoch [ 621] L = 37.952266, acc = 0.865000\n",
      "epoch [ 622] L = 37.949099, acc = 0.865000\n",
      "epoch [ 623] L = 37.945918, acc = 0.865000\n",
      "epoch [ 624] L = 37.942721, acc = 0.865000\n",
      "epoch [ 625] L = 37.939510, acc = 0.865000\n",
      "epoch [ 626] L = 37.936284, acc = 0.865000\n",
      "epoch [ 627] L = 37.933042, acc = 0.865000\n",
      "epoch [ 628] L = 37.929785, acc = 0.865000\n",
      "epoch [ 629] L = 37.926513, acc = 0.865000\n",
      "epoch [ 630] L = 37.923226, acc = 0.865000\n",
      "epoch [ 631] L = 37.919923, acc = 0.865000\n",
      "epoch [ 632] L = 37.916604, acc = 0.865000\n",
      "epoch [ 633] L = 37.913269, acc = 0.865000\n",
      "epoch [ 634] L = 37.909919, acc = 0.865000\n",
      "epoch [ 635] L = 37.906553, acc = 0.865000\n",
      "epoch [ 636] L = 37.903171, acc = 0.865000\n",
      "epoch [ 637] L = 37.899772, acc = 0.865000\n",
      "epoch [ 638] L = 37.896358, acc = 0.865000\n",
      "epoch [ 639] L = 37.892926, acc = 0.865000\n",
      "epoch [ 640] L = 37.889479, acc = 0.865000\n",
      "epoch [ 641] L = 37.886015, acc = 0.865000\n",
      "epoch [ 642] L = 37.882534, acc = 0.865000\n",
      "epoch [ 643] L = 37.879037, acc = 0.865000\n",
      "epoch [ 644] L = 37.875522, acc = 0.865000\n",
      "epoch [ 645] L = 37.871991, acc = 0.865000\n",
      "epoch [ 646] L = 37.868442, acc = 0.865000\n",
      "epoch [ 647] L = 37.864876, acc = 0.865000\n",
      "epoch [ 648] L = 37.861293, acc = 0.865000\n",
      "epoch [ 649] L = 37.857693, acc = 0.865000\n",
      "epoch [ 650] L = 37.854075, acc = 0.865000\n",
      "epoch [ 651] L = 37.850439, acc = 0.865000\n",
      "epoch [ 652] L = 37.846785, acc = 0.865000\n",
      "epoch [ 653] L = 37.843114, acc = 0.865000\n",
      "epoch [ 654] L = 37.839425, acc = 0.865000\n",
      "epoch [ 655] L = 37.835717, acc = 0.865000\n",
      "epoch [ 656] L = 37.831991, acc = 0.865000\n",
      "epoch [ 657] L = 37.828247, acc = 0.865000\n",
      "epoch [ 658] L = 37.824485, acc = 0.865000\n",
      "epoch [ 659] L = 37.820704, acc = 0.865000\n",
      "epoch [ 660] L = 37.816904, acc = 0.865000\n",
      "epoch [ 661] L = 37.813085, acc = 0.865000\n",
      "epoch [ 662] L = 37.809248, acc = 0.865000\n",
      "epoch [ 663] L = 37.805391, acc = 0.865000\n",
      "epoch [ 664] L = 37.801516, acc = 0.865000\n",
      "epoch [ 665] L = 37.797621, acc = 0.865000\n",
      "epoch [ 666] L = 37.793706, acc = 0.865000\n",
      "epoch [ 667] L = 37.789772, acc = 0.865000\n",
      "epoch [ 668] L = 37.785819, acc = 0.865000\n",
      "epoch [ 669] L = 37.781846, acc = 0.865000\n",
      "epoch [ 670] L = 37.777852, acc = 0.865000\n",
      "epoch [ 671] L = 37.773839, acc = 0.865000\n",
      "epoch [ 672] L = 37.769806, acc = 0.865000\n",
      "epoch [ 673] L = 37.765752, acc = 0.865000\n",
      "epoch [ 674] L = 37.761678, acc = 0.865000\n",
      "epoch [ 675] L = 37.757584, acc = 0.865000\n",
      "epoch [ 676] L = 37.753469, acc = 0.865000\n",
      "epoch [ 677] L = 37.749333, acc = 0.865000\n",
      "epoch [ 678] L = 37.745176, acc = 0.865000\n",
      "epoch [ 679] L = 37.740999, acc = 0.865000\n",
      "epoch [ 680] L = 37.736800, acc = 0.865000\n",
      "epoch [ 681] L = 37.732580, acc = 0.865000\n",
      "epoch [ 682] L = 37.728338, acc = 0.865000\n",
      "epoch [ 683] L = 37.724075, acc = 0.865000\n",
      "epoch [ 684] L = 37.719791, acc = 0.865000\n",
      "epoch [ 685] L = 37.715484, acc = 0.865000\n",
      "epoch [ 686] L = 37.711156, acc = 0.865000\n",
      "epoch [ 687] L = 37.706806, acc = 0.865000\n",
      "epoch [ 688] L = 37.702433, acc = 0.865000\n",
      "epoch [ 689] L = 37.698038, acc = 0.865000\n",
      "epoch [ 690] L = 37.693621, acc = 0.865000\n",
      "epoch [ 691] L = 37.689181, acc = 0.865000\n",
      "epoch [ 692] L = 37.684719, acc = 0.865000\n",
      "epoch [ 693] L = 37.680233, acc = 0.865000\n",
      "epoch [ 694] L = 37.675725, acc = 0.865000\n",
      "epoch [ 695] L = 37.671193, acc = 0.865000\n",
      "epoch [ 696] L = 37.666639, acc = 0.865000\n",
      "epoch [ 697] L = 37.662061, acc = 0.865000\n",
      "epoch [ 698] L = 37.657459, acc = 0.865000\n",
      "epoch [ 699] L = 37.652834, acc = 0.865000\n",
      "epoch [ 700] L = 37.648185, acc = 0.865000\n",
      "epoch [ 701] L = 37.643512, acc = 0.865000\n",
      "epoch [ 702] L = 37.638815, acc = 0.865000\n",
      "epoch [ 703] L = 37.634093, acc = 0.865000\n",
      "epoch [ 704] L = 37.629348, acc = 0.865000\n",
      "epoch [ 705] L = 37.624578, acc = 0.865000\n",
      "epoch [ 706] L = 37.619783, acc = 0.865000\n",
      "epoch [ 707] L = 37.614964, acc = 0.865000\n",
      "epoch [ 708] L = 37.610119, acc = 0.860000\n",
      "epoch [ 709] L = 37.605250, acc = 0.860000\n",
      "epoch [ 710] L = 37.600355, acc = 0.860000\n",
      "epoch [ 711] L = 37.595435, acc = 0.860000\n",
      "epoch [ 712] L = 37.590490, acc = 0.860000\n",
      "epoch [ 713] L = 37.585519, acc = 0.860000\n",
      "epoch [ 714] L = 37.580522, acc = 0.860000\n",
      "epoch [ 715] L = 37.575499, acc = 0.860000\n",
      "epoch [ 716] L = 37.570450, acc = 0.860000\n",
      "epoch [ 717] L = 37.565375, acc = 0.860000\n",
      "epoch [ 718] L = 37.560274, acc = 0.860000\n",
      "epoch [ 719] L = 37.555146, acc = 0.860000\n",
      "epoch [ 720] L = 37.549992, acc = 0.860000\n",
      "epoch [ 721] L = 37.544811, acc = 0.860000\n",
      "epoch [ 722] L = 37.539602, acc = 0.860000\n",
      "epoch [ 723] L = 37.534367, acc = 0.860000\n",
      "epoch [ 724] L = 37.529105, acc = 0.860000\n",
      "epoch [ 725] L = 37.523815, acc = 0.860000\n",
      "epoch [ 726] L = 37.518497, acc = 0.860000\n",
      "epoch [ 727] L = 37.513152, acc = 0.860000\n",
      "epoch [ 728] L = 37.507780, acc = 0.860000\n",
      "epoch [ 729] L = 37.502379, acc = 0.860000\n",
      "epoch [ 730] L = 37.496950, acc = 0.860000\n",
      "epoch [ 731] L = 37.491493, acc = 0.860000\n",
      "epoch [ 732] L = 37.486007, acc = 0.860000\n",
      "epoch [ 733] L = 37.480493, acc = 0.860000\n",
      "epoch [ 734] L = 37.474950, acc = 0.860000\n",
      "epoch [ 735] L = 37.469378, acc = 0.860000\n",
      "epoch [ 736] L = 37.463778, acc = 0.860000\n",
      "epoch [ 737] L = 37.458148, acc = 0.860000\n",
      "epoch [ 738] L = 37.452488, acc = 0.860000\n",
      "epoch [ 739] L = 37.446800, acc = 0.860000\n",
      "epoch [ 740] L = 37.441081, acc = 0.860000\n",
      "epoch [ 741] L = 37.435333, acc = 0.860000\n",
      "epoch [ 742] L = 37.429555, acc = 0.860000\n",
      "epoch [ 743] L = 37.423747, acc = 0.860000\n",
      "epoch [ 744] L = 37.417909, acc = 0.860000\n",
      "epoch [ 745] L = 37.412040, acc = 0.860000\n",
      "epoch [ 746] L = 37.406141, acc = 0.860000\n",
      "epoch [ 747] L = 37.400211, acc = 0.860000\n",
      "epoch [ 748] L = 37.394250, acc = 0.860000\n",
      "epoch [ 749] L = 37.388258, acc = 0.860000\n",
      "epoch [ 750] L = 37.382235, acc = 0.860000\n",
      "epoch [ 751] L = 37.376181, acc = 0.860000\n",
      "epoch [ 752] L = 37.370095, acc = 0.860000\n",
      "epoch [ 753] L = 37.363978, acc = 0.860000\n",
      "epoch [ 754] L = 37.357829, acc = 0.860000\n",
      "epoch [ 755] L = 37.351648, acc = 0.860000\n",
      "epoch [ 756] L = 37.345435, acc = 0.860000\n",
      "epoch [ 757] L = 37.339189, acc = 0.860000\n",
      "epoch [ 758] L = 37.332912, acc = 0.860000\n",
      "epoch [ 759] L = 37.326601, acc = 0.860000\n",
      "epoch [ 760] L = 37.320259, acc = 0.860000\n",
      "epoch [ 761] L = 37.313883, acc = 0.860000\n",
      "epoch [ 762] L = 37.307474, acc = 0.860000\n",
      "epoch [ 763] L = 37.301032, acc = 0.860000\n",
      "epoch [ 764] L = 37.294557, acc = 0.860000\n",
      "epoch [ 765] L = 37.288048, acc = 0.860000\n",
      "epoch [ 766] L = 37.281506, acc = 0.860000\n",
      "epoch [ 767] L = 37.274930, acc = 0.860000\n",
      "epoch [ 768] L = 37.268320, acc = 0.860000\n",
      "epoch [ 769] L = 37.261676, acc = 0.860000\n",
      "epoch [ 770] L = 37.254998, acc = 0.860000\n",
      "epoch [ 771] L = 37.248285, acc = 0.860000\n",
      "epoch [ 772] L = 37.241538, acc = 0.860000\n",
      "epoch [ 773] L = 37.234756, acc = 0.860000\n",
      "epoch [ 774] L = 37.227940, acc = 0.860000\n",
      "epoch [ 775] L = 37.221088, acc = 0.860000\n",
      "epoch [ 776] L = 37.214201, acc = 0.860000\n",
      "epoch [ 777] L = 37.207279, acc = 0.860000\n",
      "epoch [ 778] L = 37.200322, acc = 0.860000\n",
      "epoch [ 779] L = 37.193329, acc = 0.860000\n",
      "epoch [ 780] L = 37.186300, acc = 0.860000\n",
      "epoch [ 781] L = 37.179235, acc = 0.860000\n",
      "epoch [ 782] L = 37.172134, acc = 0.860000\n",
      "epoch [ 783] L = 37.164997, acc = 0.860000\n",
      "epoch [ 784] L = 37.157824, acc = 0.860000\n",
      "epoch [ 785] L = 37.150614, acc = 0.860000\n",
      "epoch [ 786] L = 37.143368, acc = 0.860000\n",
      "epoch [ 787] L = 37.136085, acc = 0.860000\n",
      "epoch [ 788] L = 37.128764, acc = 0.860000\n",
      "epoch [ 789] L = 37.121407, acc = 0.860000\n",
      "epoch [ 790] L = 37.114013, acc = 0.860000\n",
      "epoch [ 791] L = 37.106581, acc = 0.860000\n",
      "epoch [ 792] L = 37.099111, acc = 0.860000\n",
      "epoch [ 793] L = 37.091604, acc = 0.860000\n",
      "epoch [ 794] L = 37.084059, acc = 0.860000\n",
      "epoch [ 795] L = 37.076476, acc = 0.860000\n",
      "epoch [ 796] L = 37.068855, acc = 0.860000\n",
      "epoch [ 797] L = 37.061196, acc = 0.860000\n",
      "epoch [ 798] L = 37.053498, acc = 0.860000\n",
      "epoch [ 799] L = 37.045762, acc = 0.860000\n",
      "epoch [ 800] L = 37.037987, acc = 0.860000\n",
      "epoch [ 801] L = 37.030173, acc = 0.860000\n",
      "epoch [ 802] L = 37.022320, acc = 0.860000\n",
      "epoch [ 803] L = 37.014429, acc = 0.860000\n",
      "epoch [ 804] L = 37.006497, acc = 0.860000\n",
      "epoch [ 805] L = 36.998527, acc = 0.860000\n",
      "epoch [ 806] L = 36.990517, acc = 0.860000\n",
      "epoch [ 807] L = 36.982467, acc = 0.860000\n",
      "epoch [ 808] L = 36.974378, acc = 0.860000\n",
      "epoch [ 809] L = 36.966248, acc = 0.860000\n",
      "epoch [ 810] L = 36.958078, acc = 0.860000\n",
      "epoch [ 811] L = 36.949869, acc = 0.860000\n",
      "epoch [ 812] L = 36.941618, acc = 0.860000\n",
      "epoch [ 813] L = 36.933328, acc = 0.860000\n",
      "epoch [ 814] L = 36.924996, acc = 0.860000\n",
      "epoch [ 815] L = 36.916624, acc = 0.860000\n",
      "epoch [ 816] L = 36.908211, acc = 0.860000\n",
      "epoch [ 817] L = 36.899757, acc = 0.860000\n",
      "epoch [ 818] L = 36.891262, acc = 0.860000\n",
      "epoch [ 819] L = 36.882726, acc = 0.860000\n",
      "epoch [ 820] L = 36.874148, acc = 0.860000\n",
      "epoch [ 821] L = 36.865528, acc = 0.860000\n",
      "epoch [ 822] L = 36.856867, acc = 0.860000\n",
      "epoch [ 823] L = 36.848164, acc = 0.860000\n",
      "epoch [ 824] L = 36.839420, acc = 0.860000\n",
      "epoch [ 825] L = 36.830633, acc = 0.860000\n",
      "epoch [ 826] L = 36.821804, acc = 0.860000\n",
      "epoch [ 827] L = 36.812932, acc = 0.865000\n",
      "epoch [ 828] L = 36.804019, acc = 0.865000\n",
      "epoch [ 829] L = 36.795062, acc = 0.865000\n",
      "epoch [ 830] L = 36.786064, acc = 0.865000\n",
      "epoch [ 831] L = 36.777022, acc = 0.865000\n",
      "epoch [ 832] L = 36.767937, acc = 0.865000\n",
      "epoch [ 833] L = 36.758810, acc = 0.865000\n",
      "epoch [ 834] L = 36.749639, acc = 0.865000\n",
      "epoch [ 835] L = 36.740425, acc = 0.865000\n",
      "epoch [ 836] L = 36.731168, acc = 0.865000\n",
      "epoch [ 837] L = 36.721867, acc = 0.865000\n",
      "epoch [ 838] L = 36.712522, acc = 0.865000\n",
      "epoch [ 839] L = 36.703134, acc = 0.865000\n",
      "epoch [ 840] L = 36.693702, acc = 0.865000\n",
      "epoch [ 841] L = 36.684227, acc = 0.865000\n",
      "epoch [ 842] L = 36.674707, acc = 0.865000\n",
      "epoch [ 843] L = 36.665143, acc = 0.865000\n",
      "epoch [ 844] L = 36.655535, acc = 0.865000\n",
      "epoch [ 845] L = 36.645883, acc = 0.865000\n",
      "epoch [ 846] L = 36.636186, acc = 0.865000\n",
      "epoch [ 847] L = 36.626444, acc = 0.865000\n",
      "epoch [ 848] L = 36.616658, acc = 0.865000\n",
      "epoch [ 849] L = 36.606828, acc = 0.865000\n",
      "epoch [ 850] L = 36.596952, acc = 0.865000\n",
      "epoch [ 851] L = 36.587032, acc = 0.865000\n",
      "epoch [ 852] L = 36.577066, acc = 0.865000\n",
      "epoch [ 853] L = 36.567056, acc = 0.865000\n",
      "epoch [ 854] L = 36.557000, acc = 0.865000\n",
      "epoch [ 855] L = 36.546899, acc = 0.865000\n",
      "epoch [ 856] L = 36.536753, acc = 0.865000\n",
      "epoch [ 857] L = 36.526561, acc = 0.865000\n",
      "epoch [ 858] L = 36.516323, acc = 0.865000\n",
      "epoch [ 859] L = 36.506040, acc = 0.865000\n",
      "epoch [ 860] L = 36.495712, acc = 0.865000\n",
      "epoch [ 861] L = 36.485337, acc = 0.865000\n",
      "epoch [ 862] L = 36.474917, acc = 0.865000\n",
      "epoch [ 863] L = 36.464450, acc = 0.865000\n",
      "epoch [ 864] L = 36.453938, acc = 0.865000\n",
      "epoch [ 865] L = 36.443379, acc = 0.865000\n",
      "epoch [ 866] L = 36.432775, acc = 0.865000\n",
      "epoch [ 867] L = 36.422124, acc = 0.865000\n",
      "epoch [ 868] L = 36.411427, acc = 0.865000\n",
      "epoch [ 869] L = 36.400683, acc = 0.865000\n",
      "epoch [ 870] L = 36.389893, acc = 0.865000\n",
      "epoch [ 871] L = 36.379056, acc = 0.865000\n",
      "epoch [ 872] L = 36.368173, acc = 0.870000\n",
      "epoch [ 873] L = 36.357243, acc = 0.870000\n",
      "epoch [ 874] L = 36.346267, acc = 0.870000\n",
      "epoch [ 875] L = 36.335244, acc = 0.870000\n",
      "epoch [ 876] L = 36.324173, acc = 0.870000\n",
      "epoch [ 877] L = 36.313056, acc = 0.870000\n",
      "epoch [ 878] L = 36.301892, acc = 0.870000\n",
      "epoch [ 879] L = 36.290681, acc = 0.870000\n",
      "epoch [ 880] L = 36.279423, acc = 0.870000\n",
      "epoch [ 881] L = 36.268118, acc = 0.870000\n",
      "epoch [ 882] L = 36.256766, acc = 0.870000\n",
      "epoch [ 883] L = 36.245366, acc = 0.870000\n",
      "epoch [ 884] L = 36.233920, acc = 0.870000\n",
      "epoch [ 885] L = 36.222426, acc = 0.870000\n",
      "epoch [ 886] L = 36.210884, acc = 0.870000\n",
      "epoch [ 887] L = 36.199296, acc = 0.870000\n",
      "epoch [ 888] L = 36.187660, acc = 0.870000\n",
      "epoch [ 889] L = 36.175976, acc = 0.870000\n",
      "epoch [ 890] L = 36.164245, acc = 0.870000\n",
      "epoch [ 891] L = 36.152467, acc = 0.870000\n",
      "epoch [ 892] L = 36.140640, acc = 0.870000\n",
      "epoch [ 893] L = 36.128767, acc = 0.870000\n",
      "epoch [ 894] L = 36.116846, acc = 0.870000\n",
      "epoch [ 895] L = 36.104877, acc = 0.870000\n",
      "epoch [ 896] L = 36.092860, acc = 0.870000\n",
      "epoch [ 897] L = 36.080796, acc = 0.870000\n",
      "epoch [ 898] L = 36.068684, acc = 0.870000\n",
      "epoch [ 899] L = 36.056524, acc = 0.875000\n",
      "epoch [ 900] L = 36.044317, acc = 0.875000\n",
      "epoch [ 901] L = 36.032062, acc = 0.875000\n",
      "epoch [ 902] L = 36.019759, acc = 0.875000\n",
      "epoch [ 903] L = 36.007408, acc = 0.875000\n",
      "epoch [ 904] L = 35.995010, acc = 0.875000\n",
      "epoch [ 905] L = 35.982564, acc = 0.875000\n",
      "epoch [ 906] L = 35.970070, acc = 0.875000\n",
      "epoch [ 907] L = 35.957528, acc = 0.875000\n",
      "epoch [ 908] L = 35.944938, acc = 0.875000\n",
      "epoch [ 909] L = 35.932301, acc = 0.875000\n",
      "epoch [ 910] L = 35.919616, acc = 0.875000\n",
      "epoch [ 911] L = 35.906883, acc = 0.875000\n",
      "epoch [ 912] L = 35.894102, acc = 0.875000\n",
      "epoch [ 913] L = 35.881274, acc = 0.875000\n",
      "epoch [ 914] L = 35.868397, acc = 0.875000\n",
      "epoch [ 915] L = 35.855473, acc = 0.875000\n",
      "epoch [ 916] L = 35.842501, acc = 0.875000\n",
      "epoch [ 917] L = 35.829481, acc = 0.875000\n",
      "epoch [ 918] L = 35.816414, acc = 0.875000\n",
      "epoch [ 919] L = 35.803299, acc = 0.875000\n",
      "epoch [ 920] L = 35.790136, acc = 0.875000\n",
      "epoch [ 921] L = 35.776926, acc = 0.875000\n",
      "epoch [ 922] L = 35.763668, acc = 0.875000\n",
      "epoch [ 923] L = 35.750362, acc = 0.875000\n",
      "epoch [ 924] L = 35.737009, acc = 0.875000\n",
      "epoch [ 925] L = 35.723608, acc = 0.875000\n",
      "epoch [ 926] L = 35.710159, acc = 0.875000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [ 927] L = 35.696664, acc = 0.875000\n",
      "epoch [ 928] L = 35.683120, acc = 0.875000\n",
      "epoch [ 929] L = 35.669529, acc = 0.875000\n",
      "epoch [ 930] L = 35.655891, acc = 0.875000\n",
      "epoch [ 931] L = 35.642206, acc = 0.875000\n",
      "epoch [ 932] L = 35.628473, acc = 0.875000\n",
      "epoch [ 933] L = 35.614693, acc = 0.875000\n",
      "epoch [ 934] L = 35.600866, acc = 0.875000\n",
      "epoch [ 935] L = 35.586991, acc = 0.875000\n",
      "epoch [ 936] L = 35.573069, acc = 0.875000\n",
      "epoch [ 937] L = 35.559101, acc = 0.875000\n",
      "epoch [ 938] L = 35.545085, acc = 0.875000\n",
      "epoch [ 939] L = 35.531022, acc = 0.875000\n",
      "epoch [ 940] L = 35.516913, acc = 0.875000\n",
      "epoch [ 941] L = 35.502757, acc = 0.875000\n",
      "epoch [ 942] L = 35.488553, acc = 0.875000\n",
      "epoch [ 943] L = 35.474304, acc = 0.875000\n",
      "epoch [ 944] L = 35.460007, acc = 0.875000\n",
      "epoch [ 945] L = 35.445664, acc = 0.875000\n",
      "epoch [ 946] L = 35.431274, acc = 0.875000\n",
      "epoch [ 947] L = 35.416838, acc = 0.875000\n",
      "epoch [ 948] L = 35.402356, acc = 0.875000\n",
      "epoch [ 949] L = 35.387827, acc = 0.875000\n",
      "epoch [ 950] L = 35.373252, acc = 0.875000\n",
      "epoch [ 951] L = 35.358631, acc = 0.875000\n",
      "epoch [ 952] L = 35.343964, acc = 0.875000\n",
      "epoch [ 953] L = 35.329250, acc = 0.875000\n",
      "epoch [ 954] L = 35.314491, acc = 0.875000\n",
      "epoch [ 955] L = 35.299686, acc = 0.875000\n",
      "epoch [ 956] L = 35.284836, acc = 0.875000\n",
      "epoch [ 957] L = 35.269939, acc = 0.875000\n",
      "epoch [ 958] L = 35.254998, acc = 0.875000\n",
      "epoch [ 959] L = 35.240010, acc = 0.875000\n",
      "epoch [ 960] L = 35.224978, acc = 0.875000\n",
      "epoch [ 961] L = 35.209900, acc = 0.875000\n",
      "epoch [ 962] L = 35.194776, acc = 0.875000\n",
      "epoch [ 963] L = 35.179608, acc = 0.875000\n",
      "epoch [ 964] L = 35.164395, acc = 0.875000\n",
      "epoch [ 965] L = 35.149137, acc = 0.875000\n",
      "epoch [ 966] L = 35.133834, acc = 0.875000\n",
      "epoch [ 967] L = 35.118487, acc = 0.875000\n",
      "epoch [ 968] L = 35.103094, acc = 0.875000\n",
      "epoch [ 969] L = 35.087658, acc = 0.875000\n",
      "epoch [ 970] L = 35.072177, acc = 0.875000\n",
      "epoch [ 971] L = 35.056652, acc = 0.875000\n",
      "epoch [ 972] L = 35.041083, acc = 0.875000\n",
      "epoch [ 973] L = 35.025470, acc = 0.875000\n",
      "epoch [ 974] L = 35.009813, acc = 0.875000\n",
      "epoch [ 975] L = 34.994112, acc = 0.875000\n",
      "epoch [ 976] L = 34.978367, acc = 0.875000\n",
      "epoch [ 977] L = 34.962579, acc = 0.875000\n",
      "epoch [ 978] L = 34.946748, acc = 0.875000\n",
      "epoch [ 979] L = 34.930873, acc = 0.875000\n",
      "epoch [ 980] L = 34.914956, acc = 0.875000\n",
      "epoch [ 981] L = 34.898995, acc = 0.875000\n",
      "epoch [ 982] L = 34.882991, acc = 0.875000\n",
      "epoch [ 983] L = 34.866945, acc = 0.875000\n",
      "epoch [ 984] L = 34.850856, acc = 0.875000\n",
      "epoch [ 985] L = 34.834724, acc = 0.875000\n",
      "epoch [ 986] L = 34.818550, acc = 0.875000\n",
      "epoch [ 987] L = 34.802334, acc = 0.875000\n",
      "epoch [ 988] L = 34.786076, acc = 0.875000\n",
      "epoch [ 989] L = 34.769776, acc = 0.875000\n",
      "epoch [ 990] L = 34.753434, acc = 0.875000\n",
      "epoch [ 991] L = 34.737051, acc = 0.875000\n",
      "epoch [ 992] L = 34.720626, acc = 0.875000\n",
      "epoch [ 993] L = 34.704160, acc = 0.875000\n",
      "epoch [ 994] L = 34.687652, acc = 0.875000\n",
      "epoch [ 995] L = 34.671104, acc = 0.875000\n",
      "epoch [ 996] L = 34.654514, acc = 0.875000\n",
      "epoch [ 997] L = 34.637884, acc = 0.875000\n",
      "epoch [ 998] L = 34.621214, acc = 0.875000\n",
      "epoch [ 999] L = 34.604502, acc = 0.875000\n",
      "epoch [1000] L = 34.587751, acc = 0.875000\n",
      "epoch [1001] L = 34.570959, acc = 0.875000\n",
      "epoch [1002] L = 34.554128, acc = 0.875000\n",
      "epoch [1003] L = 34.537257, acc = 0.875000\n",
      "epoch [1004] L = 34.520346, acc = 0.875000\n",
      "epoch [1005] L = 34.503395, acc = 0.875000\n",
      "epoch [1006] L = 34.486406, acc = 0.875000\n",
      "epoch [1007] L = 34.469377, acc = 0.875000\n",
      "epoch [1008] L = 34.452309, acc = 0.875000\n",
      "epoch [1009] L = 34.435203, acc = 0.875000\n",
      "epoch [1010] L = 34.418057, acc = 0.875000\n",
      "epoch [1011] L = 34.400874, acc = 0.875000\n",
      "epoch [1012] L = 34.383652, acc = 0.875000\n",
      "epoch [1013] L = 34.366392, acc = 0.875000\n",
      "epoch [1014] L = 34.349094, acc = 0.875000\n",
      "epoch [1015] L = 34.331758, acc = 0.875000\n",
      "epoch [1016] L = 34.314384, acc = 0.880000\n",
      "epoch [1017] L = 34.296974, acc = 0.880000\n",
      "epoch [1018] L = 34.279526, acc = 0.880000\n",
      "epoch [1019] L = 34.262041, acc = 0.880000\n",
      "epoch [1020] L = 34.244519, acc = 0.880000\n",
      "epoch [1021] L = 34.226960, acc = 0.880000\n",
      "epoch [1022] L = 34.209365, acc = 0.880000\n",
      "epoch [1023] L = 34.191733, acc = 0.880000\n",
      "epoch [1024] L = 34.174066, acc = 0.880000\n",
      "epoch [1025] L = 34.156362, acc = 0.880000\n",
      "epoch [1026] L = 34.138623, acc = 0.880000\n",
      "epoch [1027] L = 34.120848, acc = 0.880000\n",
      "epoch [1028] L = 34.103038, acc = 0.880000\n",
      "epoch [1029] L = 34.085193, acc = 0.880000\n",
      "epoch [1030] L = 34.067312, acc = 0.880000\n",
      "epoch [1031] L = 34.049397, acc = 0.880000\n",
      "epoch [1032] L = 34.031447, acc = 0.880000\n",
      "epoch [1033] L = 34.013463, acc = 0.880000\n",
      "epoch [1034] L = 33.995445, acc = 0.880000\n",
      "epoch [1035] L = 33.977392, acc = 0.880000\n",
      "epoch [1036] L = 33.959306, acc = 0.880000\n",
      "epoch [1037] L = 33.941186, acc = 0.880000\n",
      "epoch [1038] L = 33.923032, acc = 0.880000\n",
      "epoch [1039] L = 33.904846, acc = 0.880000\n",
      "epoch [1040] L = 33.886626, acc = 0.880000\n",
      "epoch [1041] L = 33.868374, acc = 0.880000\n",
      "epoch [1042] L = 33.850089, acc = 0.880000\n",
      "epoch [1043] L = 33.831771, acc = 0.880000\n",
      "epoch [1044] L = 33.813421, acc = 0.880000\n",
      "epoch [1045] L = 33.795039, acc = 0.880000\n",
      "epoch [1046] L = 33.776626, acc = 0.880000\n",
      "epoch [1047] L = 33.758181, acc = 0.880000\n",
      "epoch [1048] L = 33.739704, acc = 0.880000\n",
      "epoch [1049] L = 33.721196, acc = 0.880000\n",
      "epoch [1050] L = 33.702657, acc = 0.880000\n",
      "epoch [1051] L = 33.684088, acc = 0.885000\n",
      "epoch [1052] L = 33.665488, acc = 0.885000\n",
      "epoch [1053] L = 33.646857, acc = 0.885000\n",
      "epoch [1054] L = 33.628196, acc = 0.885000\n",
      "epoch [1055] L = 33.609506, acc = 0.885000\n",
      "epoch [1056] L = 33.590785, acc = 0.885000\n",
      "epoch [1057] L = 33.572035, acc = 0.885000\n",
      "epoch [1058] L = 33.553256, acc = 0.885000\n",
      "epoch [1059] L = 33.534448, acc = 0.885000\n",
      "epoch [1060] L = 33.515611, acc = 0.885000\n",
      "epoch [1061] L = 33.496745, acc = 0.885000\n",
      "epoch [1062] L = 33.477850, acc = 0.885000\n",
      "epoch [1063] L = 33.458928, acc = 0.885000\n",
      "epoch [1064] L = 33.439977, acc = 0.885000\n",
      "epoch [1065] L = 33.420999, acc = 0.885000\n",
      "epoch [1066] L = 33.401992, acc = 0.890000\n",
      "epoch [1067] L = 33.382959, acc = 0.890000\n",
      "epoch [1068] L = 33.363898, acc = 0.890000\n",
      "epoch [1069] L = 33.344811, acc = 0.890000\n",
      "epoch [1070] L = 33.325697, acc = 0.890000\n",
      "epoch [1071] L = 33.306556, acc = 0.890000\n",
      "epoch [1072] L = 33.287389, acc = 0.890000\n",
      "epoch [1073] L = 33.268195, acc = 0.890000\n",
      "epoch [1074] L = 33.248976, acc = 0.890000\n",
      "epoch [1075] L = 33.229732, acc = 0.890000\n",
      "epoch [1076] L = 33.210461, acc = 0.890000\n",
      "epoch [1077] L = 33.191166, acc = 0.890000\n",
      "epoch [1078] L = 33.171846, acc = 0.890000\n",
      "epoch [1079] L = 33.152501, acc = 0.895000\n",
      "epoch [1080] L = 33.133131, acc = 0.895000\n",
      "epoch [1081] L = 33.113737, acc = 0.895000\n",
      "epoch [1082] L = 33.094319, acc = 0.895000\n",
      "epoch [1083] L = 33.074877, acc = 0.895000\n",
      "epoch [1084] L = 33.055411, acc = 0.895000\n",
      "epoch [1085] L = 33.035922, acc = 0.895000\n",
      "epoch [1086] L = 33.016410, acc = 0.895000\n",
      "epoch [1087] L = 32.996874, acc = 0.895000\n",
      "epoch [1088] L = 32.977316, acc = 0.895000\n",
      "epoch [1089] L = 32.957735, acc = 0.895000\n",
      "epoch [1090] L = 32.938132, acc = 0.895000\n",
      "epoch [1091] L = 32.918507, acc = 0.895000\n",
      "epoch [1092] L = 32.898860, acc = 0.895000\n",
      "epoch [1093] L = 32.879191, acc = 0.895000\n",
      "epoch [1094] L = 32.859501, acc = 0.895000\n",
      "epoch [1095] L = 32.839790, acc = 0.895000\n",
      "epoch [1096] L = 32.820057, acc = 0.895000\n",
      "epoch [1097] L = 32.800304, acc = 0.895000\n",
      "epoch [1098] L = 32.780530, acc = 0.895000\n",
      "epoch [1099] L = 32.760736, acc = 0.895000\n",
      "epoch [1100] L = 32.740922, acc = 0.895000\n",
      "epoch [1101] L = 32.721087, acc = 0.895000\n",
      "epoch [1102] L = 32.701233, acc = 0.895000\n",
      "epoch [1103] L = 32.681360, acc = 0.895000\n",
      "epoch [1104] L = 32.661467, acc = 0.895000\n",
      "epoch [1105] L = 32.641555, acc = 0.895000\n",
      "epoch [1106] L = 32.621625, acc = 0.895000\n",
      "epoch [1107] L = 32.601676, acc = 0.895000\n",
      "epoch [1108] L = 32.581708, acc = 0.895000\n",
      "epoch [1109] L = 32.561723, acc = 0.895000\n",
      "epoch [1110] L = 32.541719, acc = 0.895000\n",
      "epoch [1111] L = 32.521698, acc = 0.895000\n",
      "epoch [1112] L = 32.501659, acc = 0.895000\n",
      "epoch [1113] L = 32.481603, acc = 0.895000\n",
      "epoch [1114] L = 32.461530, acc = 0.895000\n",
      "epoch [1115] L = 32.441440, acc = 0.895000\n",
      "epoch [1116] L = 32.421334, acc = 0.895000\n",
      "epoch [1117] L = 32.401211, acc = 0.895000\n",
      "epoch [1118] L = 32.381072, acc = 0.895000\n",
      "epoch [1119] L = 32.360917, acc = 0.895000\n",
      "epoch [1120] L = 32.340746, acc = 0.895000\n",
      "epoch [1121] L = 32.320560, acc = 0.895000\n",
      "epoch [1122] L = 32.300358, acc = 0.895000\n",
      "epoch [1123] L = 32.280141, acc = 0.895000\n",
      "epoch [1124] L = 32.259910, acc = 0.895000\n",
      "epoch [1125] L = 32.239664, acc = 0.895000\n",
      "epoch [1126] L = 32.219403, acc = 0.895000\n",
      "epoch [1127] L = 32.199128, acc = 0.895000\n",
      "epoch [1128] L = 32.178839, acc = 0.895000\n",
      "epoch [1129] L = 32.158537, acc = 0.895000\n",
      "epoch [1130] L = 32.138220, acc = 0.895000\n",
      "epoch [1131] L = 32.117891, acc = 0.895000\n",
      "epoch [1132] L = 32.097548, acc = 0.895000\n",
      "epoch [1133] L = 32.077192, acc = 0.895000\n",
      "epoch [1134] L = 32.056824, acc = 0.895000\n",
      "epoch [1135] L = 32.036443, acc = 0.895000\n",
      "epoch [1136] L = 32.016050, acc = 0.895000\n",
      "epoch [1137] L = 31.995644, acc = 0.895000\n",
      "epoch [1138] L = 31.975227, acc = 0.895000\n",
      "epoch [1139] L = 31.954798, acc = 0.895000\n",
      "epoch [1140] L = 31.934358, acc = 0.895000\n",
      "epoch [1141] L = 31.913906, acc = 0.895000\n",
      "epoch [1142] L = 31.893444, acc = 0.895000\n",
      "epoch [1143] L = 31.872970, acc = 0.895000\n",
      "epoch [1144] L = 31.852486, acc = 0.895000\n",
      "epoch [1145] L = 31.831992, acc = 0.895000\n",
      "epoch [1146] L = 31.811487, acc = 0.895000\n",
      "epoch [1147] L = 31.790972, acc = 0.895000\n",
      "epoch [1148] L = 31.770448, acc = 0.895000\n",
      "epoch [1149] L = 31.749914, acc = 0.895000\n",
      "epoch [1150] L = 31.729371, acc = 0.895000\n",
      "epoch [1151] L = 31.708818, acc = 0.895000\n",
      "epoch [1152] L = 31.688256, acc = 0.895000\n",
      "epoch [1153] L = 31.667686, acc = 0.895000\n",
      "epoch [1154] L = 31.647107, acc = 0.895000\n",
      "epoch [1155] L = 31.626520, acc = 0.895000\n",
      "epoch [1156] L = 31.605925, acc = 0.895000\n",
      "epoch [1157] L = 31.585321, acc = 0.900000\n",
      "epoch [1158] L = 31.564710, acc = 0.900000\n",
      "epoch [1159] L = 31.544092, acc = 0.900000\n",
      "epoch [1160] L = 31.523466, acc = 0.900000\n",
      "epoch [1161] L = 31.502833, acc = 0.900000\n",
      "epoch [1162] L = 31.482192, acc = 0.900000\n",
      "epoch [1163] L = 31.461546, acc = 0.900000\n",
      "epoch [1164] L = 31.440892, acc = 0.900000\n",
      "epoch [1165] L = 31.420233, acc = 0.900000\n",
      "epoch [1166] L = 31.399567, acc = 0.900000\n",
      "epoch [1167] L = 31.378895, acc = 0.900000\n",
      "epoch [1168] L = 31.358217, acc = 0.900000\n",
      "epoch [1169] L = 31.337534, acc = 0.900000\n",
      "epoch [1170] L = 31.316845, acc = 0.900000\n",
      "epoch [1171] L = 31.296152, acc = 0.900000\n",
      "epoch [1172] L = 31.275453, acc = 0.900000\n",
      "epoch [1173] L = 31.254749, acc = 0.900000\n",
      "epoch [1174] L = 31.234041, acc = 0.900000\n",
      "epoch [1175] L = 31.213328, acc = 0.900000\n",
      "epoch [1176] L = 31.192612, acc = 0.900000\n",
      "epoch [1177] L = 31.171891, acc = 0.900000\n",
      "epoch [1178] L = 31.151166, acc = 0.900000\n",
      "epoch [1179] L = 31.130438, acc = 0.900000\n",
      "epoch [1180] L = 31.109706, acc = 0.900000\n",
      "epoch [1181] L = 31.088971, acc = 0.900000\n",
      "epoch [1182] L = 31.068232, acc = 0.900000\n",
      "epoch [1183] L = 31.047491, acc = 0.900000\n",
      "epoch [1184] L = 31.026747, acc = 0.900000\n",
      "epoch [1185] L = 31.006001, acc = 0.900000\n",
      "epoch [1186] L = 30.985252, acc = 0.900000\n",
      "epoch [1187] L = 30.964501, acc = 0.900000\n",
      "epoch [1188] L = 30.943748, acc = 0.900000\n",
      "epoch [1189] L = 30.922993, acc = 0.900000\n",
      "epoch [1190] L = 30.902236, acc = 0.900000\n",
      "epoch [1191] L = 30.881479, acc = 0.900000\n",
      "epoch [1192] L = 30.860719, acc = 0.900000\n",
      "epoch [1193] L = 30.839959, acc = 0.900000\n",
      "epoch [1194] L = 30.819198, acc = 0.900000\n",
      "epoch [1195] L = 30.798436, acc = 0.900000\n",
      "epoch [1196] L = 30.777673, acc = 0.900000\n",
      "epoch [1197] L = 30.756910, acc = 0.905000\n",
      "epoch [1198] L = 30.736147, acc = 0.905000\n",
      "epoch [1199] L = 30.715384, acc = 0.905000\n",
      "epoch [1200] L = 30.694621, acc = 0.905000\n",
      "epoch [1201] L = 30.673858, acc = 0.905000\n",
      "epoch [1202] L = 30.653096, acc = 0.905000\n",
      "epoch [1203] L = 30.632334, acc = 0.905000\n",
      "epoch [1204] L = 30.611573, acc = 0.905000\n",
      "epoch [1205] L = 30.590813, acc = 0.905000\n",
      "epoch [1206] L = 30.570055, acc = 0.905000\n",
      "epoch [1207] L = 30.549297, acc = 0.905000\n",
      "epoch [1208] L = 30.528541, acc = 0.905000\n",
      "epoch [1209] L = 30.507787, acc = 0.905000\n",
      "epoch [1210] L = 30.487034, acc = 0.905000\n",
      "epoch [1211] L = 30.466284, acc = 0.905000\n",
      "epoch [1212] L = 30.445536, acc = 0.905000\n",
      "epoch [1213] L = 30.424790, acc = 0.905000\n",
      "epoch [1214] L = 30.404046, acc = 0.905000\n",
      "epoch [1215] L = 30.383305, acc = 0.905000\n",
      "epoch [1216] L = 30.362567, acc = 0.905000\n",
      "epoch [1217] L = 30.341832, acc = 0.905000\n",
      "epoch [1218] L = 30.321100, acc = 0.905000\n",
      "epoch [1219] L = 30.300371, acc = 0.905000\n",
      "epoch [1220] L = 30.279646, acc = 0.905000\n",
      "epoch [1221] L = 30.258924, acc = 0.905000\n",
      "epoch [1222] L = 30.238207, acc = 0.905000\n",
      "epoch [1223] L = 30.217493, acc = 0.905000\n",
      "epoch [1224] L = 30.196783, acc = 0.905000\n",
      "epoch [1225] L = 30.176077, acc = 0.905000\n",
      "epoch [1226] L = 30.155375, acc = 0.905000\n",
      "epoch [1227] L = 30.134679, acc = 0.905000\n",
      "epoch [1228] L = 30.113987, acc = 0.905000\n",
      "epoch [1229] L = 30.093299, acc = 0.910000\n",
      "epoch [1230] L = 30.072617, acc = 0.910000\n",
      "epoch [1231] L = 30.051940, acc = 0.910000\n",
      "epoch [1232] L = 30.031268, acc = 0.910000\n",
      "epoch [1233] L = 30.010601, acc = 0.910000\n",
      "epoch [1234] L = 29.989940, acc = 0.915000\n",
      "epoch [1235] L = 29.969285, acc = 0.915000\n",
      "epoch [1236] L = 29.948636, acc = 0.915000\n",
      "epoch [1237] L = 29.927992, acc = 0.915000\n",
      "epoch [1238] L = 29.907355, acc = 0.915000\n",
      "epoch [1239] L = 29.886724, acc = 0.915000\n",
      "epoch [1240] L = 29.866100, acc = 0.915000\n",
      "epoch [1241] L = 29.845482, acc = 0.915000\n",
      "epoch [1242] L = 29.824871, acc = 0.915000\n",
      "epoch [1243] L = 29.804267, acc = 0.915000\n",
      "epoch [1244] L = 29.783670, acc = 0.915000\n",
      "epoch [1245] L = 29.763080, acc = 0.915000\n",
      "epoch [1246] L = 29.742497, acc = 0.915000\n",
      "epoch [1247] L = 29.721922, acc = 0.915000\n",
      "epoch [1248] L = 29.701354, acc = 0.915000\n",
      "epoch [1249] L = 29.680794, acc = 0.915000\n",
      "epoch [1250] L = 29.660242, acc = 0.915000\n",
      "epoch [1251] L = 29.639697, acc = 0.915000\n",
      "epoch [1252] L = 29.619161, acc = 0.915000\n",
      "epoch [1253] L = 29.598633, acc = 0.915000\n",
      "epoch [1254] L = 29.578114, acc = 0.915000\n",
      "epoch [1255] L = 29.557603, acc = 0.915000\n",
      "epoch [1256] L = 29.537100, acc = 0.915000\n",
      "epoch [1257] L = 29.516607, acc = 0.915000\n",
      "epoch [1258] L = 29.496122, acc = 0.915000\n",
      "epoch [1259] L = 29.475646, acc = 0.915000\n",
      "epoch [1260] L = 29.455180, acc = 0.915000\n",
      "epoch [1261] L = 29.434722, acc = 0.915000\n",
      "epoch [1262] L = 29.414274, acc = 0.915000\n",
      "epoch [1263] L = 29.393836, acc = 0.915000\n",
      "epoch [1264] L = 29.373407, acc = 0.915000\n",
      "epoch [1265] L = 29.352988, acc = 0.915000\n",
      "epoch [1266] L = 29.332579, acc = 0.915000\n",
      "epoch [1267] L = 29.312180, acc = 0.915000\n",
      "epoch [1268] L = 29.291792, acc = 0.915000\n",
      "epoch [1269] L = 29.271413, acc = 0.915000\n",
      "epoch [1270] L = 29.251045, acc = 0.915000\n",
      "epoch [1271] L = 29.230687, acc = 0.915000\n",
      "epoch [1272] L = 29.210340, acc = 0.915000\n",
      "epoch [1273] L = 29.190004, acc = 0.915000\n",
      "epoch [1274] L = 29.169678, acc = 0.915000\n",
      "epoch [1275] L = 29.149364, acc = 0.915000\n",
      "epoch [1276] L = 29.129060, acc = 0.915000\n",
      "epoch [1277] L = 29.108768, acc = 0.915000\n",
      "epoch [1278] L = 29.088487, acc = 0.915000\n",
      "epoch [1279] L = 29.068218, acc = 0.915000\n",
      "epoch [1280] L = 29.047960, acc = 0.915000\n",
      "epoch [1281] L = 29.027714, acc = 0.915000\n",
      "epoch [1282] L = 29.007479, acc = 0.915000\n",
      "epoch [1283] L = 28.987257, acc = 0.915000\n",
      "epoch [1284] L = 28.967046, acc = 0.915000\n",
      "epoch [1285] L = 28.946848, acc = 0.915000\n",
      "epoch [1286] L = 28.926662, acc = 0.915000\n",
      "epoch [1287] L = 28.906488, acc = 0.915000\n",
      "epoch [1288] L = 28.886326, acc = 0.915000\n",
      "epoch [1289] L = 28.866177, acc = 0.915000\n",
      "epoch [1290] L = 28.846041, acc = 0.915000\n",
      "epoch [1291] L = 28.825917, acc = 0.915000\n",
      "epoch [1292] L = 28.805807, acc = 0.915000\n",
      "epoch [1293] L = 28.785709, acc = 0.915000\n",
      "epoch [1294] L = 28.765624, acc = 0.915000\n",
      "epoch [1295] L = 28.745553, acc = 0.915000\n",
      "epoch [1296] L = 28.725494, acc = 0.915000\n",
      "epoch [1297] L = 28.705449, acc = 0.915000\n",
      "epoch [1298] L = 28.685418, acc = 0.915000\n",
      "epoch [1299] L = 28.665400, acc = 0.915000\n",
      "epoch [1300] L = 28.645396, acc = 0.915000\n",
      "epoch [1301] L = 28.625405, acc = 0.915000\n",
      "epoch [1302] L = 28.605429, acc = 0.915000\n",
      "epoch [1303] L = 28.585466, acc = 0.915000\n",
      "epoch [1304] L = 28.565518, acc = 0.915000\n",
      "epoch [1305] L = 28.545583, acc = 0.915000\n",
      "epoch [1306] L = 28.525663, acc = 0.915000\n",
      "epoch [1307] L = 28.505757, acc = 0.915000\n",
      "epoch [1308] L = 28.485866, acc = 0.915000\n",
      "epoch [1309] L = 28.465989, acc = 0.915000\n",
      "epoch [1310] L = 28.446126, acc = 0.915000\n",
      "epoch [1311] L = 28.426279, acc = 0.915000\n",
      "epoch [1312] L = 28.406446, acc = 0.915000\n",
      "epoch [1313] L = 28.386628, acc = 0.915000\n",
      "epoch [1314] L = 28.366825, acc = 0.915000\n",
      "epoch [1315] L = 28.347037, acc = 0.915000\n",
      "epoch [1316] L = 28.327264, acc = 0.915000\n",
      "epoch [1317] L = 28.307507, acc = 0.915000\n",
      "epoch [1318] L = 28.287764, acc = 0.915000\n",
      "epoch [1319] L = 28.268038, acc = 0.915000\n",
      "epoch [1320] L = 28.248326, acc = 0.915000\n",
      "epoch [1321] L = 28.228631, acc = 0.915000\n",
      "epoch [1322] L = 28.208951, acc = 0.915000\n",
      "epoch [1323] L = 28.189286, acc = 0.915000\n",
      "epoch [1324] L = 28.169638, acc = 0.915000\n",
      "epoch [1325] L = 28.150005, acc = 0.915000\n",
      "epoch [1326] L = 28.130388, acc = 0.915000\n",
      "epoch [1327] L = 28.110788, acc = 0.915000\n",
      "epoch [1328] L = 28.091204, acc = 0.915000\n",
      "epoch [1329] L = 28.071635, acc = 0.915000\n",
      "epoch [1330] L = 28.052084, acc = 0.915000\n",
      "epoch [1331] L = 28.032548, acc = 0.915000\n",
      "epoch [1332] L = 28.013029, acc = 0.915000\n",
      "epoch [1333] L = 27.993527, acc = 0.915000\n",
      "epoch [1334] L = 27.974041, acc = 0.915000\n",
      "epoch [1335] L = 27.954572, acc = 0.915000\n",
      "epoch [1336] L = 27.935119, acc = 0.915000\n",
      "epoch [1337] L = 27.915684, acc = 0.915000\n",
      "epoch [1338] L = 27.896265, acc = 0.915000\n",
      "epoch [1339] L = 27.876863, acc = 0.915000\n",
      "epoch [1340] L = 27.857479, acc = 0.915000\n",
      "epoch [1341] L = 27.838111, acc = 0.915000\n",
      "epoch [1342] L = 27.818761, acc = 0.915000\n",
      "epoch [1343] L = 27.799428, acc = 0.915000\n",
      "epoch [1344] L = 27.780113, acc = 0.915000\n",
      "epoch [1345] L = 27.760814, acc = 0.915000\n",
      "epoch [1346] L = 27.741534, acc = 0.915000\n",
      "epoch [1347] L = 27.722270, acc = 0.915000\n",
      "epoch [1348] L = 27.703025, acc = 0.915000\n",
      "epoch [1349] L = 27.683797, acc = 0.915000\n",
      "epoch [1350] L = 27.664587, acc = 0.915000\n",
      "epoch [1351] L = 27.645395, acc = 0.915000\n",
      "epoch [1352] L = 27.626220, acc = 0.915000\n",
      "epoch [1353] L = 27.607064, acc = 0.915000\n",
      "epoch [1354] L = 27.587925, acc = 0.915000\n",
      "epoch [1355] L = 27.568805, acc = 0.915000\n",
      "epoch [1356] L = 27.549703, acc = 0.915000\n",
      "epoch [1357] L = 27.530619, acc = 0.915000\n",
      "epoch [1358] L = 27.511553, acc = 0.915000\n",
      "epoch [1359] L = 27.492505, acc = 0.915000\n",
      "epoch [1360] L = 27.473476, acc = 0.915000\n",
      "epoch [1361] L = 27.454465, acc = 0.915000\n",
      "epoch [1362] L = 27.435473, acc = 0.915000\n",
      "epoch [1363] L = 27.416500, acc = 0.915000\n",
      "epoch [1364] L = 27.397545, acc = 0.915000\n",
      "epoch [1365] L = 27.378608, acc = 0.915000\n",
      "epoch [1366] L = 27.359691, acc = 0.915000\n",
      "epoch [1367] L = 27.340792, acc = 0.915000\n",
      "epoch [1368] L = 27.321912, acc = 0.915000\n",
      "epoch [1369] L = 27.303051, acc = 0.915000\n",
      "epoch [1370] L = 27.284209, acc = 0.915000\n",
      "epoch [1371] L = 27.265386, acc = 0.915000\n",
      "epoch [1372] L = 27.246582, acc = 0.915000\n",
      "epoch [1373] L = 27.227797, acc = 0.915000\n",
      "epoch [1374] L = 27.209031, acc = 0.915000\n",
      "epoch [1375] L = 27.190285, acc = 0.915000\n",
      "epoch [1376] L = 27.171558, acc = 0.915000\n",
      "epoch [1377] L = 27.152850, acc = 0.915000\n",
      "epoch [1378] L = 27.134161, acc = 0.915000\n",
      "epoch [1379] L = 27.115492, acc = 0.915000\n",
      "epoch [1380] L = 27.096842, acc = 0.915000\n",
      "epoch [1381] L = 27.078212, acc = 0.915000\n",
      "epoch [1382] L = 27.059602, acc = 0.915000\n",
      "epoch [1383] L = 27.041011, acc = 0.915000\n",
      "epoch [1384] L = 27.022439, acc = 0.915000\n",
      "epoch [1385] L = 27.003888, acc = 0.915000\n",
      "epoch [1386] L = 26.985356, acc = 0.915000\n",
      "epoch [1387] L = 26.966844, acc = 0.915000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [1388] L = 26.948352, acc = 0.920000\n",
      "epoch [1389] L = 26.929880, acc = 0.920000\n",
      "epoch [1390] L = 26.911427, acc = 0.920000\n",
      "epoch [1391] L = 26.892995, acc = 0.920000\n",
      "epoch [1392] L = 26.874582, acc = 0.920000\n",
      "epoch [1393] L = 26.856190, acc = 0.920000\n",
      "epoch [1394] L = 26.837818, acc = 0.920000\n",
      "epoch [1395] L = 26.819466, acc = 0.920000\n",
      "epoch [1396] L = 26.801134, acc = 0.920000\n",
      "epoch [1397] L = 26.782823, acc = 0.920000\n",
      "epoch [1398] L = 26.764531, acc = 0.920000\n",
      "epoch [1399] L = 26.746260, acc = 0.925000\n",
      "epoch [1400] L = 26.728010, acc = 0.925000\n",
      "epoch [1401] L = 26.709779, acc = 0.925000\n",
      "epoch [1402] L = 26.691569, acc = 0.925000\n",
      "epoch [1403] L = 26.673380, acc = 0.925000\n",
      "epoch [1404] L = 26.655211, acc = 0.925000\n",
      "epoch [1405] L = 26.637063, acc = 0.925000\n",
      "epoch [1406] L = 26.618935, acc = 0.925000\n",
      "epoch [1407] L = 26.600828, acc = 0.925000\n",
      "epoch [1408] L = 26.582741, acc = 0.925000\n",
      "epoch [1409] L = 26.564675, acc = 0.925000\n",
      "epoch [1410] L = 26.546630, acc = 0.925000\n",
      "epoch [1411] L = 26.528606, acc = 0.925000\n",
      "epoch [1412] L = 26.510602, acc = 0.925000\n",
      "epoch [1413] L = 26.492619, acc = 0.925000\n",
      "epoch [1414] L = 26.474657, acc = 0.925000\n",
      "epoch [1415] L = 26.456716, acc = 0.925000\n",
      "epoch [1416] L = 26.438796, acc = 0.925000\n",
      "epoch [1417] L = 26.420897, acc = 0.925000\n",
      "epoch [1418] L = 26.403018, acc = 0.925000\n",
      "epoch [1419] L = 26.385161, acc = 0.925000\n",
      "epoch [1420] L = 26.367325, acc = 0.925000\n",
      "epoch [1421] L = 26.349510, acc = 0.925000\n",
      "epoch [1422] L = 26.331715, acc = 0.925000\n",
      "epoch [1423] L = 26.313942, acc = 0.925000\n",
      "epoch [1424] L = 26.296191, acc = 0.930000\n",
      "epoch [1425] L = 26.278460, acc = 0.930000\n",
      "epoch [1426] L = 26.260750, acc = 0.930000\n",
      "epoch [1427] L = 26.243062, acc = 0.930000\n",
      "epoch [1428] L = 26.225395, acc = 0.930000\n",
      "epoch [1429] L = 26.207749, acc = 0.930000\n",
      "epoch [1430] L = 26.190125, acc = 0.930000\n",
      "epoch [1431] L = 26.172522, acc = 0.930000\n",
      "epoch [1432] L = 26.154940, acc = 0.930000\n",
      "epoch [1433] L = 26.137379, acc = 0.930000\n",
      "epoch [1434] L = 26.119840, acc = 0.930000\n",
      "epoch [1435] L = 26.102323, acc = 0.930000\n",
      "epoch [1436] L = 26.084826, acc = 0.930000\n",
      "epoch [1437] L = 26.067352, acc = 0.930000\n",
      "epoch [1438] L = 26.049898, acc = 0.930000\n",
      "epoch [1439] L = 26.032467, acc = 0.930000\n",
      "epoch [1440] L = 26.015056, acc = 0.930000\n",
      "epoch [1441] L = 25.997668, acc = 0.930000\n",
      "epoch [1442] L = 25.980301, acc = 0.930000\n",
      "epoch [1443] L = 25.962955, acc = 0.930000\n",
      "epoch [1444] L = 25.945631, acc = 0.930000\n",
      "epoch [1445] L = 25.928329, acc = 0.930000\n",
      "epoch [1446] L = 25.911048, acc = 0.930000\n",
      "epoch [1447] L = 25.893789, acc = 0.930000\n",
      "epoch [1448] L = 25.876552, acc = 0.930000\n",
      "epoch [1449] L = 25.859336, acc = 0.930000\n",
      "epoch [1450] L = 25.842142, acc = 0.930000\n",
      "epoch [1451] L = 25.824970, acc = 0.930000\n",
      "epoch [1452] L = 25.807820, acc = 0.930000\n",
      "epoch [1453] L = 25.790691, acc = 0.930000\n",
      "epoch [1454] L = 25.773584, acc = 0.930000\n",
      "epoch [1455] L = 25.756499, acc = 0.930000\n",
      "epoch [1456] L = 25.739436, acc = 0.930000\n",
      "epoch [1457] L = 25.722394, acc = 0.930000\n",
      "epoch [1458] L = 25.705374, acc = 0.930000\n",
      "epoch [1459] L = 25.688377, acc = 0.930000\n",
      "epoch [1460] L = 25.671401, acc = 0.930000\n",
      "epoch [1461] L = 25.654447, acc = 0.930000\n",
      "epoch [1462] L = 25.637514, acc = 0.930000\n",
      "epoch [1463] L = 25.620604, acc = 0.930000\n",
      "epoch [1464] L = 25.603716, acc = 0.930000\n",
      "epoch [1465] L = 25.586849, acc = 0.930000\n",
      "epoch [1466] L = 25.570004, acc = 0.930000\n",
      "epoch [1467] L = 25.553182, acc = 0.930000\n",
      "epoch [1468] L = 25.536381, acc = 0.930000\n",
      "epoch [1469] L = 25.519602, acc = 0.930000\n",
      "epoch [1470] L = 25.502846, acc = 0.930000\n",
      "epoch [1471] L = 25.486111, acc = 0.930000\n",
      "epoch [1472] L = 25.469398, acc = 0.930000\n",
      "epoch [1473] L = 25.452707, acc = 0.930000\n",
      "epoch [1474] L = 25.436038, acc = 0.930000\n",
      "epoch [1475] L = 25.419391, acc = 0.930000\n",
      "epoch [1476] L = 25.402767, acc = 0.930000\n",
      "epoch [1477] L = 25.386164, acc = 0.930000\n",
      "epoch [1478] L = 25.369583, acc = 0.930000\n",
      "epoch [1479] L = 25.353024, acc = 0.930000\n",
      "epoch [1480] L = 25.336488, acc = 0.930000\n",
      "epoch [1481] L = 25.319973, acc = 0.930000\n",
      "epoch [1482] L = 25.303480, acc = 0.930000\n",
      "epoch [1483] L = 25.287010, acc = 0.930000\n",
      "epoch [1484] L = 25.270561, acc = 0.930000\n",
      "epoch [1485] L = 25.254135, acc = 0.930000\n",
      "epoch [1486] L = 25.237731, acc = 0.930000\n",
      "epoch [1487] L = 25.221348, acc = 0.930000\n",
      "epoch [1488] L = 25.204988, acc = 0.930000\n",
      "epoch [1489] L = 25.188650, acc = 0.930000\n",
      "epoch [1490] L = 25.172334, acc = 0.930000\n",
      "epoch [1491] L = 25.156040, acc = 0.930000\n",
      "epoch [1492] L = 25.139768, acc = 0.930000\n",
      "epoch [1493] L = 25.123518, acc = 0.930000\n",
      "epoch [1494] L = 25.107291, acc = 0.930000\n",
      "epoch [1495] L = 25.091085, acc = 0.930000\n",
      "epoch [1496] L = 25.074901, acc = 0.930000\n",
      "epoch [1497] L = 25.058740, acc = 0.930000\n",
      "epoch [1498] L = 25.042601, acc = 0.930000\n",
      "epoch [1499] L = 25.026483, acc = 0.930000\n",
      "epoch [1500] L = 25.010388, acc = 0.930000\n",
      "epoch [1501] L = 24.994315, acc = 0.930000\n",
      "epoch [1502] L = 24.978264, acc = 0.930000\n",
      "epoch [1503] L = 24.962235, acc = 0.930000\n",
      "epoch [1504] L = 24.946228, acc = 0.930000\n",
      "epoch [1505] L = 24.930243, acc = 0.930000\n",
      "epoch [1506] L = 24.914280, acc = 0.930000\n",
      "epoch [1507] L = 24.898340, acc = 0.930000\n",
      "epoch [1508] L = 24.882421, acc = 0.930000\n",
      "epoch [1509] L = 24.866525, acc = 0.930000\n",
      "epoch [1510] L = 24.850650, acc = 0.930000\n",
      "epoch [1511] L = 24.834798, acc = 0.930000\n",
      "epoch [1512] L = 24.818968, acc = 0.930000\n",
      "epoch [1513] L = 24.803159, acc = 0.930000\n",
      "epoch [1514] L = 24.787373, acc = 0.930000\n",
      "epoch [1515] L = 24.771609, acc = 0.930000\n",
      "epoch [1516] L = 24.755867, acc = 0.930000\n",
      "epoch [1517] L = 24.740146, acc = 0.930000\n",
      "epoch [1518] L = 24.724448, acc = 0.930000\n",
      "epoch [1519] L = 24.708772, acc = 0.930000\n",
      "epoch [1520] L = 24.693118, acc = 0.930000\n",
      "epoch [1521] L = 24.677486, acc = 0.930000\n",
      "epoch [1522] L = 24.661876, acc = 0.930000\n",
      "epoch [1523] L = 24.646288, acc = 0.930000\n",
      "epoch [1524] L = 24.630722, acc = 0.930000\n",
      "epoch [1525] L = 24.615177, acc = 0.930000\n",
      "epoch [1526] L = 24.599655, acc = 0.930000\n",
      "epoch [1527] L = 24.584155, acc = 0.930000\n",
      "epoch [1528] L = 24.568677, acc = 0.930000\n",
      "epoch [1529] L = 24.553220, acc = 0.930000\n",
      "epoch [1530] L = 24.537786, acc = 0.930000\n",
      "epoch [1531] L = 24.522374, acc = 0.930000\n",
      "epoch [1532] L = 24.506983, acc = 0.930000\n",
      "epoch [1533] L = 24.491614, acc = 0.930000\n",
      "epoch [1534] L = 24.476267, acc = 0.930000\n",
      "epoch [1535] L = 24.460943, acc = 0.930000\n",
      "epoch [1536] L = 24.445640, acc = 0.930000\n",
      "epoch [1537] L = 24.430358, acc = 0.930000\n",
      "epoch [1538] L = 24.415099, acc = 0.930000\n",
      "epoch [1539] L = 24.399862, acc = 0.930000\n",
      "epoch [1540] L = 24.384646, acc = 0.930000\n",
      "epoch [1541] L = 24.369452, acc = 0.930000\n",
      "epoch [1542] L = 24.354280, acc = 0.930000\n",
      "epoch [1543] L = 24.339130, acc = 0.930000\n",
      "epoch [1544] L = 24.324001, acc = 0.930000\n",
      "epoch [1545] L = 24.308895, acc = 0.930000\n",
      "epoch [1546] L = 24.293810, acc = 0.930000\n",
      "epoch [1547] L = 24.278747, acc = 0.930000\n",
      "epoch [1548] L = 24.263705, acc = 0.930000\n",
      "epoch [1549] L = 24.248685, acc = 0.930000\n",
      "epoch [1550] L = 24.233687, acc = 0.930000\n",
      "epoch [1551] L = 24.218711, acc = 0.930000\n",
      "epoch [1552] L = 24.203756, acc = 0.930000\n",
      "epoch [1553] L = 24.188823, acc = 0.930000\n",
      "epoch [1554] L = 24.173912, acc = 0.930000\n",
      "epoch [1555] L = 24.159022, acc = 0.930000\n",
      "epoch [1556] L = 24.144154, acc = 0.930000\n",
      "epoch [1557] L = 24.129307, acc = 0.930000\n",
      "epoch [1558] L = 24.114482, acc = 0.935000\n",
      "epoch [1559] L = 24.099679, acc = 0.935000\n",
      "epoch [1560] L = 24.084897, acc = 0.935000\n",
      "epoch [1561] L = 24.070137, acc = 0.935000\n",
      "epoch [1562] L = 24.055398, acc = 0.935000\n",
      "epoch [1563] L = 24.040681, acc = 0.935000\n",
      "epoch [1564] L = 24.025985, acc = 0.935000\n",
      "epoch [1565] L = 24.011310, acc = 0.935000\n",
      "epoch [1566] L = 23.996657, acc = 0.935000\n",
      "epoch [1567] L = 23.982026, acc = 0.935000\n",
      "epoch [1568] L = 23.967416, acc = 0.935000\n",
      "epoch [1569] L = 23.952827, acc = 0.935000\n",
      "epoch [1570] L = 23.938260, acc = 0.935000\n",
      "epoch [1571] L = 23.923714, acc = 0.935000\n",
      "epoch [1572] L = 23.909189, acc = 0.935000\n",
      "epoch [1573] L = 23.894686, acc = 0.935000\n",
      "epoch [1574] L = 23.880204, acc = 0.935000\n",
      "epoch [1575] L = 23.865744, acc = 0.935000\n",
      "epoch [1576] L = 23.851304, acc = 0.935000\n",
      "epoch [1577] L = 23.836886, acc = 0.935000\n",
      "epoch [1578] L = 23.822489, acc = 0.935000\n",
      "epoch [1579] L = 23.808113, acc = 0.935000\n",
      "epoch [1580] L = 23.793759, acc = 0.935000\n",
      "epoch [1581] L = 23.779426, acc = 0.935000\n",
      "epoch [1582] L = 23.765113, acc = 0.935000\n",
      "epoch [1583] L = 23.750822, acc = 0.935000\n",
      "epoch [1584] L = 23.736552, acc = 0.935000\n",
      "epoch [1585] L = 23.722303, acc = 0.935000\n",
      "epoch [1586] L = 23.708076, acc = 0.935000\n",
      "epoch [1587] L = 23.693869, acc = 0.935000\n",
      "epoch [1588] L = 23.679683, acc = 0.940000\n",
      "epoch [1589] L = 23.665518, acc = 0.940000\n",
      "epoch [1590] L = 23.651375, acc = 0.940000\n",
      "epoch [1591] L = 23.637252, acc = 0.940000\n",
      "epoch [1592] L = 23.623150, acc = 0.940000\n",
      "epoch [1593] L = 23.609069, acc = 0.940000\n",
      "epoch [1594] L = 23.595009, acc = 0.940000\n",
      "epoch [1595] L = 23.580970, acc = 0.940000\n",
      "epoch [1596] L = 23.566952, acc = 0.940000\n",
      "epoch [1597] L = 23.552954, acc = 0.940000\n",
      "epoch [1598] L = 23.538978, acc = 0.940000\n",
      "epoch [1599] L = 23.525022, acc = 0.940000\n",
      "epoch [1600] L = 23.511087, acc = 0.940000\n",
      "epoch [1601] L = 23.497172, acc = 0.940000\n",
      "epoch [1602] L = 23.483278, acc = 0.940000\n",
      "epoch [1603] L = 23.469406, acc = 0.940000\n",
      "epoch [1604] L = 23.455553, acc = 0.940000\n",
      "epoch [1605] L = 23.441722, acc = 0.940000\n",
      "epoch [1606] L = 23.427910, acc = 0.940000\n",
      "epoch [1607] L = 23.414120, acc = 0.940000\n",
      "epoch [1608] L = 23.400350, acc = 0.940000\n",
      "epoch [1609] L = 23.386601, acc = 0.940000\n",
      "epoch [1610] L = 23.372872, acc = 0.940000\n",
      "epoch [1611] L = 23.359164, acc = 0.940000\n",
      "epoch [1612] L = 23.345476, acc = 0.940000\n",
      "epoch [1613] L = 23.331808, acc = 0.940000\n",
      "epoch [1614] L = 23.318161, acc = 0.940000\n",
      "epoch [1615] L = 23.304535, acc = 0.940000\n",
      "epoch [1616] L = 23.290929, acc = 0.940000\n",
      "epoch [1617] L = 23.277343, acc = 0.940000\n",
      "epoch [1618] L = 23.263777, acc = 0.940000\n",
      "epoch [1619] L = 23.250232, acc = 0.940000\n",
      "epoch [1620] L = 23.236707, acc = 0.940000\n",
      "epoch [1621] L = 23.223202, acc = 0.940000\n",
      "epoch [1622] L = 23.209718, acc = 0.940000\n",
      "epoch [1623] L = 23.196254, acc = 0.940000\n",
      "epoch [1624] L = 23.182809, acc = 0.940000\n",
      "epoch [1625] L = 23.169385, acc = 0.940000\n",
      "epoch [1626] L = 23.155982, acc = 0.940000\n",
      "epoch [1627] L = 23.142598, acc = 0.940000\n",
      "epoch [1628] L = 23.129234, acc = 0.940000\n",
      "epoch [1629] L = 23.115890, acc = 0.940000\n",
      "epoch [1630] L = 23.102567, acc = 0.940000\n",
      "epoch [1631] L = 23.089263, acc = 0.940000\n",
      "epoch [1632] L = 23.075979, acc = 0.940000\n",
      "epoch [1633] L = 23.062715, acc = 0.940000\n",
      "epoch [1634] L = 23.049471, acc = 0.940000\n",
      "epoch [1635] L = 23.036247, acc = 0.940000\n",
      "epoch [1636] L = 23.023043, acc = 0.940000\n",
      "epoch [1637] L = 23.009859, acc = 0.940000\n",
      "epoch [1638] L = 22.996694, acc = 0.940000\n",
      "epoch [1639] L = 22.983549, acc = 0.940000\n",
      "epoch [1640] L = 22.970424, acc = 0.940000\n",
      "epoch [1641] L = 22.957319, acc = 0.940000\n",
      "epoch [1642] L = 22.944233, acc = 0.940000\n",
      "epoch [1643] L = 22.931167, acc = 0.940000\n",
      "epoch [1644] L = 22.918120, acc = 0.940000\n",
      "epoch [1645] L = 22.905093, acc = 0.940000\n",
      "epoch [1646] L = 22.892086, acc = 0.940000\n",
      "epoch [1647] L = 22.879098, acc = 0.940000\n",
      "epoch [1648] L = 22.866130, acc = 0.940000\n",
      "epoch [1649] L = 22.853181, acc = 0.940000\n",
      "epoch [1650] L = 22.840252, acc = 0.940000\n",
      "epoch [1651] L = 22.827342, acc = 0.940000\n",
      "epoch [1652] L = 22.814451, acc = 0.940000\n",
      "epoch [1653] L = 22.801580, acc = 0.940000\n",
      "epoch [1654] L = 22.788728, acc = 0.940000\n",
      "epoch [1655] L = 22.775895, acc = 0.940000\n",
      "epoch [1656] L = 22.763082, acc = 0.940000\n",
      "epoch [1657] L = 22.750288, acc = 0.940000\n",
      "epoch [1658] L = 22.737513, acc = 0.940000\n",
      "epoch [1659] L = 22.724758, acc = 0.940000\n",
      "epoch [1660] L = 22.712021, acc = 0.940000\n",
      "epoch [1661] L = 22.699304, acc = 0.940000\n",
      "epoch [1662] L = 22.686605, acc = 0.940000\n",
      "epoch [1663] L = 22.673926, acc = 0.940000\n",
      "epoch [1664] L = 22.661266, acc = 0.940000\n",
      "epoch [1665] L = 22.648625, acc = 0.940000\n",
      "epoch [1666] L = 22.636003, acc = 0.940000\n",
      "epoch [1667] L = 22.623399, acc = 0.940000\n",
      "epoch [1668] L = 22.610815, acc = 0.940000\n",
      "epoch [1669] L = 22.598249, acc = 0.940000\n",
      "epoch [1670] L = 22.585703, acc = 0.940000\n",
      "epoch [1671] L = 22.573175, acc = 0.940000\n",
      "epoch [1672] L = 22.560666, acc = 0.940000\n",
      "epoch [1673] L = 22.548176, acc = 0.940000\n",
      "epoch [1674] L = 22.535704, acc = 0.940000\n",
      "epoch [1675] L = 22.523251, acc = 0.940000\n",
      "epoch [1676] L = 22.510817, acc = 0.940000\n",
      "epoch [1677] L = 22.498402, acc = 0.940000\n",
      "epoch [1678] L = 22.486005, acc = 0.940000\n",
      "epoch [1679] L = 22.473626, acc = 0.940000\n",
      "epoch [1680] L = 22.461267, acc = 0.940000\n",
      "epoch [1681] L = 22.448925, acc = 0.940000\n",
      "epoch [1682] L = 22.436602, acc = 0.940000\n",
      "epoch [1683] L = 22.424298, acc = 0.940000\n",
      "epoch [1684] L = 22.412012, acc = 0.940000\n",
      "epoch [1685] L = 22.399745, acc = 0.940000\n",
      "epoch [1686] L = 22.387495, acc = 0.940000\n",
      "epoch [1687] L = 22.375264, acc = 0.940000\n",
      "epoch [1688] L = 22.363052, acc = 0.940000\n",
      "epoch [1689] L = 22.350858, acc = 0.940000\n",
      "epoch [1690] L = 22.338681, acc = 0.940000\n",
      "epoch [1691] L = 22.326524, acc = 0.940000\n",
      "epoch [1692] L = 22.314384, acc = 0.940000\n",
      "epoch [1693] L = 22.302262, acc = 0.940000\n",
      "epoch [1694] L = 22.290159, acc = 0.940000\n",
      "epoch [1695] L = 22.278073, acc = 0.940000\n",
      "epoch [1696] L = 22.266006, acc = 0.940000\n",
      "epoch [1697] L = 22.253956, acc = 0.940000\n",
      "epoch [1698] L = 22.241925, acc = 0.940000\n",
      "epoch [1699] L = 22.229911, acc = 0.940000\n",
      "epoch [1700] L = 22.217916, acc = 0.940000\n",
      "epoch [1701] L = 22.205938, acc = 0.940000\n",
      "epoch [1702] L = 22.193978, acc = 0.940000\n",
      "epoch [1703] L = 22.182036, acc = 0.940000\n",
      "epoch [1704] L = 22.170112, acc = 0.940000\n",
      "epoch [1705] L = 22.158205, acc = 0.940000\n",
      "epoch [1706] L = 22.146317, acc = 0.940000\n",
      "epoch [1707] L = 22.134446, acc = 0.940000\n",
      "epoch [1708] L = 22.122592, acc = 0.940000\n",
      "epoch [1709] L = 22.110756, acc = 0.940000\n",
      "epoch [1710] L = 22.098938, acc = 0.940000\n",
      "epoch [1711] L = 22.087137, acc = 0.940000\n",
      "epoch [1712] L = 22.075354, acc = 0.940000\n",
      "epoch [1713] L = 22.063588, acc = 0.940000\n",
      "epoch [1714] L = 22.051840, acc = 0.940000\n",
      "epoch [1715] L = 22.040109, acc = 0.940000\n",
      "epoch [1716] L = 22.028396, acc = 0.940000\n",
      "epoch [1717] L = 22.016700, acc = 0.940000\n",
      "epoch [1718] L = 22.005021, acc = 0.940000\n",
      "epoch [1719] L = 21.993359, acc = 0.940000\n",
      "epoch [1720] L = 21.981715, acc = 0.940000\n",
      "epoch [1721] L = 21.970088, acc = 0.940000\n",
      "epoch [1722] L = 21.958478, acc = 0.940000\n",
      "epoch [1723] L = 21.946886, acc = 0.940000\n",
      "epoch [1724] L = 21.935310, acc = 0.940000\n",
      "epoch [1725] L = 21.923752, acc = 0.940000\n",
      "epoch [1726] L = 21.912210, acc = 0.945000\n",
      "epoch [1727] L = 21.900686, acc = 0.945000\n",
      "epoch [1728] L = 21.889178, acc = 0.945000\n",
      "epoch [1729] L = 21.877688, acc = 0.945000\n",
      "epoch [1730] L = 21.866215, acc = 0.945000\n",
      "epoch [1731] L = 21.854758, acc = 0.945000\n",
      "epoch [1732] L = 21.843318, acc = 0.945000\n",
      "epoch [1733] L = 21.831895, acc = 0.945000\n",
      "epoch [1734] L = 21.820489, acc = 0.945000\n",
      "epoch [1735] L = 21.809100, acc = 0.945000\n",
      "epoch [1736] L = 21.797727, acc = 0.945000\n",
      "epoch [1737] L = 21.786371, acc = 0.945000\n",
      "epoch [1738] L = 21.775032, acc = 0.945000\n",
      "epoch [1739] L = 21.763709, acc = 0.945000\n",
      "epoch [1740] L = 21.752403, acc = 0.945000\n",
      "epoch [1741] L = 21.741114, acc = 0.945000\n",
      "epoch [1742] L = 21.729841, acc = 0.945000\n",
      "epoch [1743] L = 21.718584, acc = 0.945000\n",
      "epoch [1744] L = 21.707344, acc = 0.945000\n",
      "epoch [1745] L = 21.696120, acc = 0.945000\n",
      "epoch [1746] L = 21.684913, acc = 0.945000\n",
      "epoch [1747] L = 21.673722, acc = 0.945000\n",
      "epoch [1748] L = 21.662548, acc = 0.945000\n",
      "epoch [1749] L = 21.651390, acc = 0.945000\n",
      "epoch [1750] L = 21.640248, acc = 0.945000\n",
      "epoch [1751] L = 21.629122, acc = 0.945000\n",
      "epoch [1752] L = 21.618012, acc = 0.945000\n",
      "epoch [1753] L = 21.606919, acc = 0.945000\n",
      "epoch [1754] L = 21.595841, acc = 0.945000\n",
      "epoch [1755] L = 21.584780, acc = 0.945000\n",
      "epoch [1756] L = 21.573735, acc = 0.945000\n",
      "epoch [1757] L = 21.562706, acc = 0.945000\n",
      "epoch [1758] L = 21.551693, acc = 0.945000\n",
      "epoch [1759] L = 21.540695, acc = 0.945000\n",
      "epoch [1760] L = 21.529714, acc = 0.945000\n",
      "epoch [1761] L = 21.518749, acc = 0.945000\n",
      "epoch [1762] L = 21.507799, acc = 0.945000\n",
      "epoch [1763] L = 21.496865, acc = 0.945000\n",
      "epoch [1764] L = 21.485947, acc = 0.945000\n",
      "epoch [1765] L = 21.475045, acc = 0.945000\n",
      "epoch [1766] L = 21.464159, acc = 0.945000\n",
      "epoch [1767] L = 21.453288, acc = 0.945000\n",
      "epoch [1768] L = 21.442433, acc = 0.945000\n",
      "epoch [1769] L = 21.431593, acc = 0.945000\n",
      "epoch [1770] L = 21.420769, acc = 0.945000\n",
      "epoch [1771] L = 21.409961, acc = 0.945000\n",
      "epoch [1772] L = 21.399168, acc = 0.945000\n",
      "epoch [1773] L = 21.388391, acc = 0.945000\n",
      "epoch [1774] L = 21.377629, acc = 0.945000\n",
      "epoch [1775] L = 21.366882, acc = 0.945000\n",
      "epoch [1776] L = 21.356151, acc = 0.945000\n",
      "epoch [1777] L = 21.345435, acc = 0.945000\n",
      "epoch [1778] L = 21.334735, acc = 0.945000\n",
      "epoch [1779] L = 21.324049, acc = 0.945000\n",
      "epoch [1780] L = 21.313379, acc = 0.945000\n",
      "epoch [1781] L = 21.302725, acc = 0.945000\n",
      "epoch [1782] L = 21.292085, acc = 0.945000\n",
      "epoch [1783] L = 21.281460, acc = 0.945000\n",
      "epoch [1784] L = 21.270851, acc = 0.945000\n",
      "epoch [1785] L = 21.260257, acc = 0.945000\n",
      "epoch [1786] L = 21.249678, acc = 0.945000\n",
      "epoch [1787] L = 21.239113, acc = 0.945000\n",
      "epoch [1788] L = 21.228564, acc = 0.945000\n",
      "epoch [1789] L = 21.218030, acc = 0.945000\n",
      "epoch [1790] L = 21.207510, acc = 0.945000\n",
      "epoch [1791] L = 21.197006, acc = 0.945000\n",
      "epoch [1792] L = 21.186516, acc = 0.945000\n",
      "epoch [1793] L = 21.176041, acc = 0.945000\n",
      "epoch [1794] L = 21.165581, acc = 0.945000\n",
      "epoch [1795] L = 21.155136, acc = 0.945000\n",
      "epoch [1796] L = 21.144705, acc = 0.945000\n",
      "epoch [1797] L = 21.134289, acc = 0.945000\n",
      "epoch [1798] L = 21.123888, acc = 0.945000\n",
      "epoch [1799] L = 21.113501, acc = 0.945000\n",
      "epoch [1800] L = 21.103129, acc = 0.945000\n",
      "epoch [1801] L = 21.092771, acc = 0.945000\n",
      "epoch [1802] L = 21.082428, acc = 0.945000\n",
      "epoch [1803] L = 21.072099, acc = 0.945000\n",
      "epoch [1804] L = 21.061785, acc = 0.945000\n",
      "epoch [1805] L = 21.051485, acc = 0.945000\n",
      "epoch [1806] L = 21.041199, acc = 0.945000\n",
      "epoch [1807] L = 21.030928, acc = 0.945000\n",
      "epoch [1808] L = 21.020671, acc = 0.945000\n",
      "epoch [1809] L = 21.010429, acc = 0.945000\n",
      "epoch [1810] L = 21.000200, acc = 0.945000\n",
      "epoch [1811] L = 20.989986, acc = 0.945000\n",
      "epoch [1812] L = 20.979786, acc = 0.945000\n",
      "epoch [1813] L = 20.969600, acc = 0.945000\n",
      "epoch [1814] L = 20.959428, acc = 0.945000\n",
      "epoch [1815] L = 20.949270, acc = 0.945000\n",
      "epoch [1816] L = 20.939127, acc = 0.945000\n",
      "epoch [1817] L = 20.928997, acc = 0.945000\n",
      "epoch [1818] L = 20.918881, acc = 0.945000\n",
      "epoch [1819] L = 20.908779, acc = 0.945000\n",
      "epoch [1820] L = 20.898691, acc = 0.945000\n",
      "epoch [1821] L = 20.888617, acc = 0.945000\n",
      "epoch [1822] L = 20.878557, acc = 0.945000\n",
      "epoch [1823] L = 20.868510, acc = 0.945000\n",
      "epoch [1824] L = 20.858478, acc = 0.945000\n",
      "epoch [1825] L = 20.848459, acc = 0.945000\n",
      "epoch [1826] L = 20.838453, acc = 0.945000\n",
      "epoch [1827] L = 20.828462, acc = 0.945000\n",
      "epoch [1828] L = 20.818484, acc = 0.945000\n",
      "epoch [1829] L = 20.808519, acc = 0.945000\n",
      "epoch [1830] L = 20.798568, acc = 0.945000\n",
      "epoch [1831] L = 20.788631, acc = 0.945000\n",
      "epoch [1832] L = 20.778707, acc = 0.945000\n",
      "epoch [1833] L = 20.768797, acc = 0.945000\n",
      "epoch [1834] L = 20.758900, acc = 0.945000\n",
      "epoch [1835] L = 20.749016, acc = 0.945000\n",
      "epoch [1836] L = 20.739146, acc = 0.945000\n",
      "epoch [1837] L = 20.729289, acc = 0.945000\n",
      "epoch [1838] L = 20.719446, acc = 0.945000\n",
      "epoch [1839] L = 20.709615, acc = 0.945000\n",
      "epoch [1840] L = 20.699798, acc = 0.945000\n",
      "epoch [1841] L = 20.689994, acc = 0.945000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [1842] L = 20.680203, acc = 0.945000\n",
      "epoch [1843] L = 20.670426, acc = 0.945000\n",
      "epoch [1844] L = 20.660661, acc = 0.945000\n",
      "epoch [1845] L = 20.650909, acc = 0.945000\n",
      "epoch [1846] L = 20.641171, acc = 0.945000\n",
      "epoch [1847] L = 20.631445, acc = 0.945000\n",
      "epoch [1848] L = 20.621733, acc = 0.945000\n",
      "epoch [1849] L = 20.612033, acc = 0.945000\n",
      "epoch [1850] L = 20.602346, acc = 0.945000\n",
      "epoch [1851] L = 20.592673, acc = 0.945000\n",
      "epoch [1852] L = 20.583012, acc = 0.945000\n",
      "epoch [1853] L = 20.573363, acc = 0.945000\n",
      "epoch [1854] L = 20.563728, acc = 0.945000\n",
      "epoch [1855] L = 20.554105, acc = 0.945000\n",
      "epoch [1856] L = 20.544495, acc = 0.945000\n",
      "epoch [1857] L = 20.534897, acc = 0.945000\n",
      "epoch [1858] L = 20.525312, acc = 0.945000\n",
      "epoch [1859] L = 20.515740, acc = 0.945000\n",
      "epoch [1860] L = 20.506181, acc = 0.945000\n",
      "epoch [1861] L = 20.496633, acc = 0.945000\n",
      "epoch [1862] L = 20.487099, acc = 0.945000\n",
      "epoch [1863] L = 20.477576, acc = 0.945000\n",
      "epoch [1864] L = 20.468067, acc = 0.945000\n",
      "epoch [1865] L = 20.458569, acc = 0.945000\n",
      "epoch [1866] L = 20.449084, acc = 0.945000\n",
      "epoch [1867] L = 20.439611, acc = 0.945000\n",
      "epoch [1868] L = 20.430151, acc = 0.945000\n",
      "epoch [1869] L = 20.420703, acc = 0.945000\n",
      "epoch [1870] L = 20.411267, acc = 0.945000\n",
      "epoch [1871] L = 20.401843, acc = 0.945000\n",
      "epoch [1872] L = 20.392432, acc = 0.945000\n",
      "epoch [1873] L = 20.383032, acc = 0.945000\n",
      "epoch [1874] L = 20.373645, acc = 0.945000\n",
      "epoch [1875] L = 20.364269, acc = 0.945000\n",
      "epoch [1876] L = 20.354906, acc = 0.945000\n",
      "epoch [1877] L = 20.345555, acc = 0.945000\n",
      "epoch [1878] L = 20.336216, acc = 0.945000\n",
      "epoch [1879] L = 20.326888, acc = 0.945000\n",
      "epoch [1880] L = 20.317573, acc = 0.945000\n",
      "epoch [1881] L = 20.308269, acc = 0.945000\n",
      "epoch [1882] L = 20.298977, acc = 0.945000\n",
      "epoch [1883] L = 20.289697, acc = 0.945000\n",
      "epoch [1884] L = 20.280429, acc = 0.945000\n",
      "epoch [1885] L = 20.271172, acc = 0.945000\n",
      "epoch [1886] L = 20.261928, acc = 0.945000\n",
      "epoch [1887] L = 20.252694, acc = 0.945000\n",
      "epoch [1888] L = 20.243473, acc = 0.945000\n",
      "epoch [1889] L = 20.234263, acc = 0.945000\n",
      "epoch [1890] L = 20.225065, acc = 0.945000\n",
      "epoch [1891] L = 20.215878, acc = 0.945000\n",
      "epoch [1892] L = 20.206703, acc = 0.945000\n",
      "epoch [1893] L = 20.197539, acc = 0.945000\n",
      "epoch [1894] L = 20.188386, acc = 0.945000\n",
      "epoch [1895] L = 20.179245, acc = 0.945000\n",
      "epoch [1896] L = 20.170116, acc = 0.945000\n",
      "epoch [1897] L = 20.160998, acc = 0.945000\n",
      "epoch [1898] L = 20.151891, acc = 0.945000\n",
      "epoch [1899] L = 20.142795, acc = 0.945000\n",
      "epoch [1900] L = 20.133710, acc = 0.945000\n",
      "epoch [1901] L = 20.124637, acc = 0.945000\n",
      "epoch [1902] L = 20.115575, acc = 0.945000\n",
      "epoch [1903] L = 20.106524, acc = 0.945000\n",
      "epoch [1904] L = 20.097485, acc = 0.945000\n",
      "epoch [1905] L = 20.088456, acc = 0.945000\n",
      "epoch [1906] L = 20.079438, acc = 0.945000\n",
      "epoch [1907] L = 20.070431, acc = 0.945000\n",
      "epoch [1908] L = 20.061436, acc = 0.945000\n",
      "epoch [1909] L = 20.052451, acc = 0.945000\n",
      "epoch [1910] L = 20.043477, acc = 0.950000\n",
      "epoch [1911] L = 20.034514, acc = 0.950000\n",
      "epoch [1912] L = 20.025562, acc = 0.950000\n",
      "epoch [1913] L = 20.016621, acc = 0.950000\n",
      "epoch [1914] L = 20.007691, acc = 0.950000\n",
      "epoch [1915] L = 19.998771, acc = 0.950000\n",
      "epoch [1916] L = 19.989862, acc = 0.950000\n",
      "epoch [1917] L = 19.980964, acc = 0.950000\n",
      "epoch [1918] L = 19.972076, acc = 0.950000\n",
      "epoch [1919] L = 19.963199, acc = 0.950000\n",
      "epoch [1920] L = 19.954333, acc = 0.950000\n",
      "epoch [1921] L = 19.945477, acc = 0.950000\n",
      "epoch [1922] L = 19.936632, acc = 0.950000\n",
      "epoch [1923] L = 19.927797, acc = 0.950000\n",
      "epoch [1924] L = 19.918973, acc = 0.950000\n",
      "epoch [1925] L = 19.910159, acc = 0.950000\n",
      "epoch [1926] L = 19.901355, acc = 0.950000\n",
      "epoch [1927] L = 19.892562, acc = 0.950000\n",
      "epoch [1928] L = 19.883779, acc = 0.950000\n",
      "epoch [1929] L = 19.875007, acc = 0.950000\n",
      "epoch [1930] L = 19.866245, acc = 0.950000\n",
      "epoch [1931] L = 19.857493, acc = 0.950000\n",
      "epoch [1932] L = 19.848751, acc = 0.950000\n",
      "epoch [1933] L = 19.840019, acc = 0.950000\n",
      "epoch [1934] L = 19.831298, acc = 0.950000\n",
      "epoch [1935] L = 19.822587, acc = 0.950000\n",
      "epoch [1936] L = 19.813885, acc = 0.950000\n",
      "epoch [1937] L = 19.805194, acc = 0.950000\n",
      "epoch [1938] L = 19.796513, acc = 0.950000\n",
      "epoch [1939] L = 19.787842, acc = 0.950000\n",
      "epoch [1940] L = 19.779181, acc = 0.950000\n",
      "epoch [1941] L = 19.770529, acc = 0.950000\n",
      "epoch [1942] L = 19.761888, acc = 0.950000\n",
      "epoch [1943] L = 19.753257, acc = 0.950000\n",
      "epoch [1944] L = 19.744635, acc = 0.950000\n",
      "epoch [1945] L = 19.736023, acc = 0.950000\n",
      "epoch [1946] L = 19.727421, acc = 0.950000\n",
      "epoch [1947] L = 19.718828, acc = 0.950000\n",
      "epoch [1948] L = 19.710246, acc = 0.950000\n",
      "epoch [1949] L = 19.701673, acc = 0.950000\n",
      "epoch [1950] L = 19.693109, acc = 0.950000\n",
      "epoch [1951] L = 19.684555, acc = 0.950000\n",
      "epoch [1952] L = 19.676011, acc = 0.950000\n",
      "epoch [1953] L = 19.667476, acc = 0.950000\n",
      "epoch [1954] L = 19.658951, acc = 0.950000\n",
      "epoch [1955] L = 19.650436, acc = 0.950000\n",
      "epoch [1956] L = 19.641929, acc = 0.950000\n",
      "epoch [1957] L = 19.633432, acc = 0.950000\n",
      "epoch [1958] L = 19.624945, acc = 0.950000\n",
      "epoch [1959] L = 19.616467, acc = 0.950000\n",
      "epoch [1960] L = 19.607998, acc = 0.950000\n",
      "epoch [1961] L = 19.599539, acc = 0.950000\n",
      "epoch [1962] L = 19.591088, acc = 0.950000\n",
      "epoch [1963] L = 19.582647, acc = 0.950000\n",
      "epoch [1964] L = 19.574216, acc = 0.950000\n",
      "epoch [1965] L = 19.565793, acc = 0.950000\n",
      "epoch [1966] L = 19.557379, acc = 0.950000\n",
      "epoch [1967] L = 19.548975, acc = 0.950000\n",
      "epoch [1968] L = 19.540580, acc = 0.950000\n",
      "epoch [1969] L = 19.532193, acc = 0.950000\n",
      "epoch [1970] L = 19.523816, acc = 0.950000\n",
      "epoch [1971] L = 19.515448, acc = 0.950000\n",
      "epoch [1972] L = 19.507088, acc = 0.950000\n",
      "epoch [1973] L = 19.498738, acc = 0.950000\n",
      "epoch [1974] L = 19.490396, acc = 0.950000\n",
      "epoch [1975] L = 19.482063, acc = 0.950000\n",
      "epoch [1976] L = 19.473739, acc = 0.950000\n",
      "epoch [1977] L = 19.465424, acc = 0.950000\n",
      "epoch [1978] L = 19.457118, acc = 0.950000\n",
      "epoch [1979] L = 19.448820, acc = 0.950000\n",
      "epoch [1980] L = 19.440531, acc = 0.950000\n",
      "epoch [1981] L = 19.432251, acc = 0.950000\n",
      "epoch [1982] L = 19.423979, acc = 0.950000\n",
      "epoch [1983] L = 19.415716, acc = 0.950000\n",
      "epoch [1984] L = 19.407461, acc = 0.950000\n",
      "epoch [1985] L = 19.399215, acc = 0.950000\n",
      "epoch [1986] L = 19.390978, acc = 0.950000\n",
      "epoch [1987] L = 19.382749, acc = 0.950000\n",
      "epoch [1988] L = 19.374528, acc = 0.950000\n",
      "epoch [1989] L = 19.366316, acc = 0.950000\n",
      "epoch [1990] L = 19.358112, acc = 0.950000\n",
      "epoch [1991] L = 19.349917, acc = 0.950000\n",
      "epoch [1992] L = 19.341730, acc = 0.950000\n",
      "epoch [1993] L = 19.333551, acc = 0.950000\n",
      "epoch [1994] L = 19.325380, acc = 0.950000\n",
      "epoch [1995] L = 19.317218, acc = 0.950000\n",
      "epoch [1996] L = 19.309064, acc = 0.950000\n",
      "epoch [1997] L = 19.300918, acc = 0.950000\n",
      "epoch [1998] L = 19.292780, acc = 0.950000\n",
      "epoch [1999] L = 19.284650, acc = 0.950000\n"
     ]
    }
   ],
   "source": [
    "# FIXME: change variable name to math\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "y_true = np.array(nn.y).astype(float)\n",
    "\n",
    "# back-propagation\n",
    "def backpropagation(n, X, y):\n",
    "    for i in range(n.n_epoch):\n",
    "        # forward to calculate each node's output\n",
    "        forward(n, X)\n",
    "        \n",
    "        # print loss, accuracy\n",
    "        L = np.sum((n.z2 - y)**2)\n",
    "        \n",
    "        y_pred = np.argmax(nn.z2, axis=1)\n",
    "        acc = accuracy_score(y_true, y_pred)\n",
    "        \n",
    "        print(\"epoch [%4d] L = %f, acc = %f\" % (i, L, acc))\n",
    "        \n",
    "        # calc weights update\n",
    "        d2 = n.z2*(1-n.z2)*(y - n.z2)\n",
    "        d1 = n.z1*(1-n.z1)*(np.dot(d2, n.W2.T))\n",
    "        \n",
    "        # update weights\n",
    "        n.W2 += n.epsilon * np.dot(n.z1.T, d2)\n",
    "        n.b2 += n.epsilon * np.sum(d2, axis=0)\n",
    "        n.W1 += n.epsilon * np.dot(X.T, d1)\n",
    "        n.b1 += n.epsilon * np.sum(d1, axis=0)\n",
    "\n",
    "nn.n_epoch = 2000\n",
    "backpropagation(nn, X, t)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAEICAYAAABcVE8dAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAAB+fElEQVR4nO2ddXgUV/fHP3dmLW4kIYQEd3d3hwItUvf2bfurv3V395a6y9tS2lKhFC/u7pYQCDFCiNva7Pz+2BCy7MYd5vM8PCSzM/ee3eyevXPuOd8jVFVFQ0NDQ+PCR6pvAzQ0NDQ06gbN4WtoaGhcJGgOX0NDQ+MiQXP4GhoaGhcJmsPX0NDQuEjQHL6GhobGRYLm8DU0qoEQYrUQ4tY6nvNGIcT6upxT48JAc/gaGrWEEOI5IcT/qjlGSyGEKoTQ1ZRdGhcvmsPXuKBoTI5RONE+gxp1hvZm02jwCCF6CyF2CSFyhRC/CiHmCSFeKnpspBAiUQjxqBDiFPCNEMIohHhPCJFc9O89IYSx6Hy3cEjRCrpt0c/fCiE+EkL8UzTfFiFEmxLnjhNCHBZCZAshPgREKTZPBJ4ArhBC5Akh9hQdXy2EeFkIsQEoAFoLIU4IIcaWuLbkncHaov+zisYZVOK8t4QQmUKI40KISdV5jTUuDjSHr9GgEUIYgD+Ab4FgYC5w2XmnNS16rAVwG/AkMBDoCfQA+gNPVWLaK4HngSAgFni5yJYmwO9FYzUBjgFDPA2gquoS4BVgnqqqvqqq9ijx8HVFdvoB8eXYMrzo/8CicTYV/T4AOFJkxxvAV0IIj18+Ghpn0Ry+RkNnIKADPlBV1aaq6u/A1vPOcQDPqqpqUVW1ELgGeEFV1dOqqqbhdN7XVWLOP1RV3aqqqh34EecXB8Bk4ICqqr+pqmoD3gNOVeE5fauq6gFVVe1F41SFeFVVv1BVVQG+AyKA8CqOpXGRoDl8jYZOMyBJdVX5SzjvnDRVVc3nXVNy5RxfdKyilHTiBYBviXGL5y6y6XxbKkJVrjmfYhtVVS0o+tG3lHM1NADN4Ws0fFKAyPPCFVHnnXO+5GsyzvDOWaKLjgHkA95nHxBCNK2kLcVzF9l0vi1l2VXacRebcIaoyhtDQ6PSaA5fo6GzCVCAu4UQOiHEdJwx+bKYCzwlhAgtirs/A5zdBN0DdBFC9BRCmIDnKmHLP0XXzijKBroXV+d8PqlAywpk4uwGrhRC6IUQfYFZJR5Lwxmyal0JOzU0PKI5fI0GjaqqVmAGcAuQBVwLLAQsZVz2ErAd2AvsA3YWHUNV1aPAC8AKIAaocAGTqqpngNnAa0A60A7YUMYlvxb9ny6E2FnGeU8DbYBMnPsNP5WYswDnpvEGIUSWEGJgRe3V0DgfoTVA0WhsCCG2AJ+qqvpNfduiodGY0Fb4Gg0eIcQIIUTTopDODUB3YEl926Wh0dhoNFWJGhc1HYBfAB8gDpilqmpK/ZqkodH40EI6GhoaGhcJWkhHQ0ND4yKhwYZ0mjRporZs2bK+zdDQ0NBoVOzYseOMqqqhnh5rsA6/ZcuWbN++vb7N0NDQ0GhUCCFK1WfSQjoaGhoaFwmaw9fQ0NC4SNAcvoaGhsZFgubwNTQ0NC4SNIevUWXMaVmcWruXvJOp9W2KhoZGBWiwWToaDRfV4WDzfR9y9MtFyCYDDouNZmN7M/Lnp9F5m+rbPA0NjVLQVvgalebgnD+I+WYJDosNW3Y+itlK8oqdbLr7g/o2DQCHzU7m/uPkJ6XVtykaGg0KzeFrVJqD789HKXBVJ1bMVuLmrkSxWOvJKidx81YxN3wmCwffw/x217NoxP0Uns6sV5s0NBoKmsPXqDSWzDzPDzhU7IX15/DPbD/C+lvexJqVhz2vEMVs5fTmgyyf8kS92aSh0ZDQHL5GpWk6oge4dBx04hMdhr3ATMaeY/Wy0t//7m8o533hqDaFrEPxZB08Uef2aGg0NDSHr1Fp+r1xG3o/LyS9c89fyBKylwFDkC/z217HouH381PoDA5/vrBO7cpPOA0e1F8lvY6ClIw6tUVDoyGiZeloVJqA9lFcuu8r9r/1C2mbDxLQMZqcmETSd8bgsNpRzM5V9tYHPsa/dQTNxvbBnJ7NoQ//JGnZdnxbhNPlv7MI7dexRu2KHNeX9O1Hi+c/i8NiI6RX2xqdS0OjMdJg9fD79u2rauJpjYP8pDTmt70OxWJze6zZuD4M/+Fx/up1G9bMPKczFgLZy8DQLx+i9ZWja8wOS0YOf3a/FfOZbBxWOwA6HxNdH5xNr+durLF5NDQaMkKIHaqq9vX0mBbS0ag2hamZSEa9x8cKks6w7/WfsZzJObfyVlWUAgub7nwfh81eY3YYg/2ZvutzOt19Kf4doggb3IVh3z6qOXsNjSK0kI5GtQns1AJVcbgdl/Q6mo3rQ8I/mz06doddIftoIkFdWtaYLabQQPq/9X/0f+v/amxMDY0LBW2Fr1FtdF5G+rx6K7K3sfiY0MvoA7zp9vAVmJoEeLxOtdkxBvnWlZkaGhc92gpfo0bofPdl+LeNZN+b8yhITidyXB+6P3YV3s2a0OW/s8jYcwx7vrn4fKHXETqwM97NmtSj1RoaFxeaw9eoMZpP7E/zif3djreYMYzM/cfZ9/rPSEY9DpudoK6tGPXLM/VgpYbGxYuWpaNRZ1gyc8nYHYtXRAiBHaPr2xwNjQuSsrJ0tBW+Rp1hDPIjYlSv+jbDDcVSlCpq8JxppKFxoaA5/AuY7JhE8hPSCO7eutSN04uZvPhU1t/yJqfW7AEBEaN7M/TLh/BpHlrfpmlo1Aqaw68BEhZt4cC7v2E+nUXUJQPp8sAsTCH152AtWXn8e+nTnNl2BMmgQzFb6XT3pfR743aEBw2cixF7oYWFg+7GfDoL1eFMKU35dycLB9/NrNj/1clqX3U4iP9zA8fnrUI2GWh/8ySnTpGGRi2hOfxqsveNn9nz4g/FGSjZRxOI/WEZl+7+AmOwf73YtO6G10jbfNApc1DolDE+8unfBHZuSfubJtaLTQ2N+PlrseUVFjt7AFVxYMsu4ORfG2k1e0Stzq+qKitnPkvyip3O944QxP++js73z6TPizfX6twaFy9aHn41sGbnsfu571zSDR0WG+Yz2Rz88M96scmSlUfSsu3F0gJnseebOfjub/ViU0MkJzYZe16h23F7gZnc2KRanz95+Y5zzh5AVbHnmznw9q/knjhV6/NrXJxoDr8aZOw+5lFSwGG2kbRkaz1YBLacfITk+c9qycitY2saLkHdW6Pz9XI7LnsbCereutbnP/n3RpeFwlmEJJG8fEetz69xcaI5/GpgCgv0rAUjBN6R9bPx59M8FEOgj9txoZOJnOgxU+uiJHrqILybhRRLPANIBj2+0eFETuxX6/MbAnwROtn9AVmg93P/ItLQqAk0h18NAju1ILBTC7cPrs7LQJf7Z9aLTUKSGPLZA8jexuKVvmTUYwjwqTURsfykNA68N5+9r80lfc8xEv7ZzOFP/yZ9V0ytzAdQcCqDXc99y/JLnmDH099QkJJeqeslvY5LNs6h7Q3j0Qf4YAj0pd1NE5m87n3Madke++EWnMpg/X/eYm7Tmfza+hr2vTkPh12pkv1trx+HpHd3+AKImjqoSmNqaJSHVnhVTQpTM/j3smfI2HMMSa9DVVUGvHcX7W+aVK92pe+O5cC7v5ETm0TEyJ50vvcyvMKDa3yeY3P/ZcMtbwGg2OygOJAMOoQsO1MdR/VizO/Pu6ykq0vW4ZMsHHQ3itmKw2JDMuqRjXqmbJhTLSG27KMJrL7qJbIPxoMQ+LaKYORPTxLcow3W7Dz+6HIzhaezUIucvOxtJGrKQEbNq1rFcOz/lrPx9neLHb8QgjELXqLpsO5Vfg4aGmUVXmkOv4bIPZ6CJT2HwK6t0JkM9W1OnWDJyGFe8yvcGo6URPYy0uuFG+n24OU1Nu+ScQ+RsnK3a3crIWg6vDuTVr1TpTHtZiu/trwKc1q2y7iGAB9mn/iJmG+WsOOpr92at8smA9N3f05A+6gqzWvNySdl1W5ko56IUT2RjRfHe0ej9tD08OsAv1YRNOnb4aJx9gCJi7d6jkOXQCm0cPTzf2p03lNr9rq3MlRVTq3bS1UXMCf/2oC90OI2rsOmEDd3Fanr97s5e3Cqgqbviq3SnAAGfx9aTB9C84n9NWevUetoDl+j1nFY3TthVQedl2fHKJsM5RaWqQ4H+Ylp2HILXI7nJ6ThMLvbaS8wk3cylYAOUUgGD2EpVcW3RXjFjdfQqEc0h69RZZpP6l8czy4Nyain5RUja3TetjdNRD7vTko2GWh7w4Qyrzvxx3rmNb+c+R1u4KewGay+6kVsRbn4of07enToOl8vwgZ2psPtl7jtQwi9Dr9WzQgd0Kmaz0hDo27QHH4jQbHasBe4523XJ8Zgf4Z89RCyyVBqi0OvsCB6PH51jc7b99X/0HREd2QvI3p/b2QvI2FDu9L/zdtLvSZtyyHWXvsKhacyUQotOCw24v/cwOqrXgIgfFg3mvTtgOx1romLbDLg3y6S5lMG4BsdzoRlbzhX+kY9kkFH5LjeTFzxpiZXodFo0DZtGziWzFw23v4OJxdsRFUcBPdsy5AvHiSkZ9v6Nq2Y/KQ01t/0Jimrdrm1OjRFBHNlwrxSi8GqQ9aheLIOnSSwYxSBnVuWee6/M5/l5J8b3GL0ssnAzJjv8YkMRbFY2ffGPPa/8yu27HwQAp/mTRj0yX+Jmjyg+BpzWhayyYDez7vGn5OGRnXRNm0bKaqqsnTcw5xcsBGH1Y6qOEjfcZTFI/5b6bzz2sQnMpS8k6ke+9racwrIiakdqYLATi1oOWNYuc4eIPdYsvtGL86QU36CM+deNhrIiUk8J0uhquQnpLHq8udJ23IIh6IQ+8NyVl/9MquueIET89dWeZNYQ6M+qJHkaCHE18AlwGlVVbt6eFwA7wOTgQLgRlVVd9bE3BcyZ7YeJvtIgpsujmK1ceSLf+j1zPVVHlt1OEhatp2M3cfwax1B9PTB1coS0XmbPM+jOFx63dYklsxc9r3xMyfmr0Pv60Wnuy+l3Y0TPN5NhA/rRtaheFSb656Dw2IjsJOzGYs5PZsTv611SzNVCq3sfvl/zkyg1XuKJRFS1+0j4Z/NDPv6kVp5fhoaNU1NrfC/BcqSYZwEtCv6dxvwSQ3Ne0GTcywZIbnHhx0WG1n7j1d5XGtOPgv63sGqy19g59Nfs+E/b/Fr62vJi0+t8pgd/28aOh9Xpy8kicAuLfCNCqvyuKVhyy/k737/x4H35pMbm0TG7li23PshG273nIff7ZEr0ft4QYkvA523iS4PzcYQ4Gyknp+QVmomTsbuWBdnD05BuuPzVpOx91jNPjkNjVqiRhy+qqprgYwyTpkOfK862QwECiEiamLuC5ng7q1x2N3DJLK3kdCBnas87q5nviXr0EnseYVOSeDcQsynM1l30+tVHrP9LZNoOXM4spcBna8XOj8vvJs3YdSvz1V5zLI49sMKCk5l4rCcS6W0F5iJ+/Ffj2qTvlFhTN32Ca2uGIEpPIjAri0Z9Mn99H7+puJz/Ns0w2FzzzoSsoQ+wMej2JnqcDiLwCqAw65w/Nc1rLv5DbY//iU5daDKqaFRkrrSw48EEkr8nlh0LKXkSUKI23DeARAdrfU8DeraiqYje3Bq9W6UQmeYQcgSel8v2t1cdemGY3P/dXGU4Ay9pK7fjy2/0LkSriRCkhj27aN0f/Ia0jYfwrtZCBGjetbKZi1A8r87UTxkLUl6HWlbDuHXsqnbY/5tmjHyx6dKHVPv502X+2dy8P3fXTKiZC8jUZMHkhub7Pa6SXodxmC/cu1VLFYWj36QzL1x2PPNCL3MwQ9+Z/gPj9NyxrByr9fQqAka1Katqqqfq6raV1XVvqGhWps5gDG/P0+XB2ZjCgtE7+9Ny1kjmLrtE4yBvlUftBY3GgPaNaftdeNoNqZ3rTl7AN+W4QgP4mOqquLdLKTK4/Z+6Wb6vXU7vq2aovfzptm4PkxZ/z5d7pvh8fkIAS0uG1ruuEe/XkLmnrjiuwTVpqAUWlg1+3l2Pv+dZ9VVDY0apq5W+ElASbGR5kXHNMpBNhro8+LNNdoFqdXlozj65SLXClhJEDawc5VW9/VBxzumceSTBdhLhGCELOHVNJjwod2qPK4Qgo53TKPjHdPcHhs9/znWXPWSMzNHdWb4jPnzxQqlZx7/eaXnOgpVZd+rc8naG8fo+c9X2W4NjYpQVw5/AXC3EOJnYACQrapqSjnXaNQSvV+6mZRVu8hPSMOeV4jO1wvZZCC4V1vmd7wB2ainw+1TndWlctlaOfWFf5tmjP79Bdbd9Dq2nAJUxUFQjzaM+uWZcguh8hPTsBda8G8bWamiqeYT+3Nl6nzSNh9C6GRCB3Ss8Otz/oZ2SRxWG4lLtpJ9JIGADlUTYdPQqAg1UnglhJgLjASaAKnAs4AeQFXVT4vSMj/EmclTANykqmqZVVVa4VXt4rArJCzcRMbuY/hEh7H/nV/IiztVnJIoe5toPqkfo2tp07UmsGbnkbp+H9ZcM+FDupSbDZQXn8rK2c+Rtf8EQpIwBPky/IfHiRjZs9Ztjf9rA2uvfcXjxi849w+GfPEgrS4fWeu2aFzYaPLIFynW7DwKT2fh2yIc2eBZ+gAg7ueVbLjtHbcer7K3kUs2fUhwt9aoqopitlZIoKwuODjnD7Y/+jmSUY/qcGAM9GP80tcJ7Oh5s9+hKPzW9joKEtJcGpfrfExcduBrfKNrVwBNVVW2PvAxhz7802OBms7HxMSVbxPar2Ot2qFx4aNV2l5k2M1W1lz3CnObzmJBnzuYGzaDQx//Ver5p9bs8djQ22Gzs+/1uex+9Ufmhs3gf/6XMC9yNke/WVyb5pfL6U0H2P74FyhmK7bsfOy5heQnprFs/CMuzrwkp1btxpqR6/a4w6Zw9MtFtW6zEIIB797FpNXvuukOSXodAR2iaNK3Q63boXFxozn8OkBVVY79bzl/9rqNeVFXsO6WN8g7ea7ISbHaKDydiUOpWru889l4xzvE/74Oh8WGPa8QW04B2x/5jJMLNno83yc6HMmDjr9qUzj+yxp2Pfk1lvQcVMVB4alMNt8zh2NzV9aIrVXh0CcLitNUi1FVrNl5nN500OM1BcnpqKr7l4HDaiN9dyz73viZg3P+qHXJivAhXZm85j0Cu7ZE0stIBh3NpwxgwrI3GsSdk8aFjRbSqQN2PPkVBz/4vTh+K2QJQ6Av03d/zuFP/+bge/OdEgReRnq/eBOd7pxe5bmsOfn8HD4TxeKu7d6kf0embv7I7XhBSjrz219fanzZE35tmjEr5ocq21kdlk1+jKQl29yO6/19GPHTky5CZ2fJPpLAX71uc5NNEDq52NEKnXP9M+y7x2g1a4THuRWrjWM/LOfY/5Yjm4x0uP0SoqcPqZKztmTlIRv16LxqR3pC4+KkrJBOXWXpXLRYMnM58O5vLo7mbHXr6iteIGP3MexFnZQUs5Vtj3yGIdCXNlePKXVMu9nKkc/+5tgPy5H0Ojrcdgltrh+HJMtY0nOcXag8OPyCRPfG3ADeESGMX/waa655hfykM1BKWKQk+Qmnyz2ntmhx2TBS1+5zS3N0WG2EDe7i8ZqADlG0mDmMk39sKL5O6GVUu+OcAFrRS7buhteJHNenWHKheHxFYemER0nfdqR4jNT1+2h7w3gGfXhfpZ9HabUUZ3YcZf9b88iJTabp8O50fXA23s2aVHp8DY3z0UI6tUzm/uMeteIdVhunNx8qdvZnUQos7H7h+1LHcygKS8c+xI4nvyJ9ZwxpWw6x+d45rLnmFQB8osI8NgwXkkRYGfnp4UO7MfvETzQb06tCz8u/XSQABcln2PXC96y98TVivlnibBNYy7S5bhwBHaPOCbYJgextpM+rt5ZZkDbs20fp9/YdBHVvjV/bSIK6tfZ4ntDJJHq4g0hYuJn0HUdcvmjs+WZivl5Cdkxi9Z5UESf/3siiEfdz/Jc1pO84yqGP/uSPbrd6lIvQ0KgsmsOvZXyah7qV4wPOEs1SwmkFyWdKHS9x0RYy9sa59Fe155tJWLiJ9N2xSDqZvm/e7qJQKSQJnY+R3s/fWKatQgiPImjnI3sb6fvabaRuPMD8Djew99WfOPb9cjbfO4c/u9+KJTO3zOuri85kYPL6D+j/7p1ETuhH66tGM2HpG3S5b2aZ10myTMfbp3Lp7i+YdfR7QvuVtkmqevzbJC3dhj3PQ9hLEpxavacKz+S8WR0ONt7xrvNvWzS/w2rHlpPPrme+qfb4Ghqaw69l/FpFEDqoM9J5aZGylwFjsL/Ha4K6tip1vJTVnjNqVMVB6rp9AHS4ZTKjf32WsMFd8IkOo+XlTjmGihT1RE8bTPSlQ5xOXxLOblYGHd6RTZBNBoK6tWbUL8/SfPIA1l7nzCs/+4VmzzeTn3Cava/+VO481UVnMtDhP1MYv/g1RvzvCcKHuKlyl0luXDKnNx7w6NhVu0LkxP5ux73CAj2qaUqyXCE9nfLISziNNSvP3R7FQfIKTU1co/poMfw6YMzvz7PuxtdJXLINSZbQ+Xkz6OP7UQrNbLjtHZfVuuxtpN8bpbfq844IRjYZ3DYfJYMOr/Cg4t+bTxpA80num5flISSJ4d8/zpltR0hashV9gA+trxyFV3iwy3l5J1MpTHEXSHVY7Rz/dU2Zz6G+sZut/DPkXgpPZ7k9JpsMDP3qYY+hobY3TGDfm78Arro3QifTfMrAKttzZvsRNv7fe6TvOFrqOcYQz4sDDY3KoDn8OsAQ4MuYP17EkpmLNTsf3+iwYiEuQ4Avu579ltzjpwjq1oo+r9xKeCkbjwBtrh3H7ufdY/ySXkfUtME1Yq8QgtD+HQntX3oRkGwylNrtSedV9UYqdcHJP9Zjyze7re4lg54+r9xC6ytHe7zOr1UEI+c+xdrrX3UeUFX0ft6M/ftldB7SWitCblwyi0c/6PGu7Sw6HxNdHpiN3WwlfcdRdD4mgnu00dI4NSqN5vArSH5SGnFzV2JJzyFyfF+ajuxZ6Q+cMcgPY5DrrX/UlIFEVWJ1KJsMjJz3DOtvedOZRqmqmMKCGPPH81V2OlXBKyyIJn3ak7blkEvlqM7b6FF4rL7JPprA1gc+JmXVHoTA4+ayw2rDml1Q5jjR0wZz1enfSdt8CNlkoEnf9lVWBS1MzWD9rW+X2pxe5+eFalPodPelyF4Gfg6fAUKgOhx4hQczbuErmvaORqXQ8vArQMI/m1l1xQuoigOHxYbOx0TE6F6M/v35OhMXyz6SwNrrXyVjt7O7UpP+Hen++NX4tQwnoFMLhBDYzVZO/rmevPhUQvt3rNKXUmXISzjN4hH/xZyeDQ4V1aHSfHJ/Rs59GknXcETXClLS+aPLzViz88uUhtb5ejHif08QXUN3SmVx7McVbPjP2yhWGzjcbdL5edHrmRtod/NECpLT+bv/nSglv6SEwLtZCLNP/NRgBe406gctD78aKBYra6552S0rJmXlLk78sobWV3m+/a9JbHmF/DP0XiwZucUOK23TQTb85y1mx/0IOFewi4bdj2K2YC90at4E92jDhGVv1Fphj29UGLNifyBl5S7yE9No0q8jQV1a1spc1eHgnD+cK/qyFjcCjE38aT6l8vselaXgVIbT2Z+3D1MSpdDKsbn/krxiB5Je5yplDaCq2HIKOLV6D83G9K5lizUuFDSHXw6nNx7weNyebyb2h2V14vCP/7La6RxKOCzV4cCWU8DfA+4i68AJVIdrKqE9r5D0nUfZ98bP9Hr2hlqzTUgSzcb2qbXxa4Iz2w57To0tiQrm1Cwy98YR0qtdrdpz8s8N4KFXcTFCoKoqGTtjnL/KkkfBNVAxn8muHSM1Lki0tMxyELIMpSwMywtb2AstHP91DUc+X0jOseQq25Abl+JR9sCebyZzX5zTGXhYvSqFVmK/W1bleS8Ugrq1RngoRjsfxWxlzyu1n1LqsCsewzjgrP4VOglKOHjPzt4p/BY+tHLpqBoXN5rDL4ewwV085l7rfExl9pVN23qYec1ms/7WN9nywMf82e0Wttz/UamZLWXRpG97dL6ldKIqZ7jS1CMvJjrfcxmyh2pnN1SVrAPHa92e6KmDPB6XvY00HdYd1UMjdSThUsuh8zHR+b6Z+ERqrUA1Ko7m8MtB0smM+eMF9H5e6Hy9kIx6ZC8jra4cRfT0IR6vcSgKK6Y+ibVIulcpsKCYrRz9ahGJi7ZU2oaoSwbhGx3mItEg5PL/dLLJQJtrxlZ6vgsNv1YRTPz3LUJ6t0NIEsKg8/j6CVkipHf7WrfHt0U4vV+8CdnL4NQ9kiRkLyOd776UgI7RHm3TeZto/5/JhA7qTOSEfoz46Un6vHJLrduqcWGhZelUEFtuAfF/rMeSkUuzsb3LrIZNXb+P5VOewJbrnuIXPX0IY/54odLzW7Pz2PX898TNXYmQBIGdW3B64wF3meAidL5e+LeLZPKa99CXdndwEaJYbUg6ma0PfcqRzxe6bMbrfExM3foxgZ1a1IktWQdPEPfLalS7g5YzhxHSqx2Z+4/z98C7XOxCCLwjQ5h9XMvI0SgfreNVHZP8705WznwWW467w282vi8Tlrxe7Tms2Xn81u56l6YewqDDp1kIra8ZS2j/jjSfPAAhSaxZEcvSvw6Rn2+ha49mzLymJyGhPtW2oSFgL7Sw/fEviP1mCfZCK01HdGfgnHtL7Xx1FtXh4MC7v7H/nV+xZOQS2r8j/d+5kyZ9an+FXx5xP69kw+3vIIRAVRx4NQ1m3D+vEND+wsm5z8u1sOzvQ+zenohfgImJ0zrTrVez+jbrgkBz+HWMvdDC3PCZbtWTOh8Tgz6+n7bXjauReXJPnGLLvXNIWrodyaCj9dVj6P/WHej9vIvP+enr7axaehSrxRkXliTw8jHwygfTCAxq/Cv/pRMe4dTqPThs5+QO9AE+zDj0Ld5Ng8u4smFztqpW7+tFUPfWF1RVbX6ehafuX0hOthm7zblYMRhlZlzVg0mXll5lrlExtBaHdYzOy8iwbx5B9jIWSxXrfE2EDepSo2mcfi2bMnbBy9xgWcp1uf8w5LMHXJx9bo6ZlYuPFDt7cErdW8x2lv19qMbsqC8SFm0mefkOF2cPzi/cI58uqCeragadyUD4kK4XpITCikVHyM2xFDt7AKtF4fef9lBYUHptQkMnK7OQXVsTOB6bXqXkjLpAy8OvJVrOHE5I73bEfrcU85lsmk8eSPOJ/apchl8VEuOz0OllbDbXTB27zcHh/amlXNV42HTXBx6Pq1Y7Z3YcxaEozk3aC8xhNnb2bE/CZnXPRJJ1EvFxmXTsWrsN5WsaVVX5+ZsdrFh8BL1exuFQaRLqw8PPjyUo2Lv8AeoQzeHXIn6tIuj13I1Vvt6hKNiy89EH+FRpsy64iTd2u3tappAETZtVX863PslPTKPwlLta51nSd8XynWECkkFH2+vH0/+d/0PvU3chrLz4VPa9NY+0zYcI7NyCrg9dTnApDVcuNoKCvUHgllKsKA78A8ruxdAQ2bL+BKuWxmC3OYrvWlKScpjz2hqeeaP01O36QHP4DRBVVdn/5jz2vPoTSqEVnbeRns9eT+d7Z1RqtRoe4U/b9k2IOZzm4vj1eomJ0zvXhuk1jqqqpK7fx+n1+/FqGkzLWcPR+3njsCtl3i0VJjubkTssNo79sIzcuGQmLn+rTmzOOhTPwkF3Yy+0oNoUMnbFEj9/LWMWvEyz0RXrKFYdClLSif1hGQWJZ4gY1YuoqYMalLbRhGmd2LsrySXUKEmCiMgAmkUF1KNlnsnLsbBtUzwF+Va69IigZZsQl8eX/n0Yi8U1rOhwqJw8nkl6Wn6DSpDQHH4NkBObxI4nv+LU6t0YmwTQ9aEraHfjhCqHEg5+8Du7X/yhuLrWarWx88mv0fmY6HDrlEqNde/jI/lyzkb2bE9CSAIfXwM33zmI6FYNf0PTYbOzfOqTnN6wH8Xs1Afa+sDHTPj3LUJ6tcM7MoTcWA8VzJJwqWRVzDZObzpI1qH4Okm53PbIZ9hyC4urn1WHA3uBhU3/9y4zj5TevrImOLV2L8unPI6qOFDMVmK+XUpgp2gmrX63wTRLb985jKtv7svcb3YgCYGiOIiMDuT+J0bWt2luHNiTwvuvrEZFRbE7+HPeXvoPbsGt9w4u/nwX5nved5BkQWFhOZIedYyWpVNN8k6m8meP/2DPLSxOj9R5m+h83wz6vFy1wpi54TMxp2W5HfduHsoVJ3+u0piFBVYKC+0EBXs1mpj2wQ//YPtjX7jmpAO+LZsy69j/SN9xlCVjHsJhV1AKLeh8vRACp7M9D72/N8O+e4wWpRTL1ST/C5jqsQZD6GSuSf/TZWO9JlEdDuZFXeHWmEb2MtDruRvp9vAVtTJvVbFa7CTEZ+LrZyQ8ouE1eLHZFO654VcKC1ydttGk447/DqX3AGea7C/f72Tp34dcNqEBfP2MzPl2FlIFiiRrEi1LpxbZ98bP2AvMLhIG9gIzB979DWu2e7u68lAdDo/OHigzZl0eXt4GgkO8G42zB4j5eombswcwp2WRfSSBJn07MPv4j/R97Va6PDib4d8/RvDssTg87Hc4rHYCO9dNQZUh0PMtvKSTkWuxZ0HWoZMeaz+UQivHflxRa/NWFYNRR5v2oQ3S2QMcPXjao3SJxWxn3cpjxb9PvqwLAYFeGAzO950kCQxGmVvuHlTnzr48tJBONUldt9+j9olk1JN16CRhAysXKxeShG+rCPKOp7g9FtC+eZXtbJSUdfdZ9Jgx2J/O98wAYNvGeBaePkxPISNQOPvVJox6mo3vS0C7unn9uvx3Fjuf+hp7ydaVJgOtrxlTnKZbG0h6udR0QNlQAS0hDXdKWR+VfJ19/Yy8/P4lrFkey/7dKYSE+jBuSgeatwjyfHE90rC+fhoh/u0iwcOqWTFbyU84zdGvF5O+O7ZSY/Z76w5kb9d4q+xlpN9bd1TL1oaKxWIvynQ4yulTucXH2944we11AGd/14DzKmlVVWXuNzvIl4zsHD6FjLBIFEnCpjeQ2a0no+Y9XevP4yyd753htN1kQB/gg2wyEDmhHwM/uKdW5/Vv1xyf5qFu70edj4n2/6nc3o8GtOsU5nHNYTTpGDqqjcsxL28DE6d35qFnx3DTnQMbpLMHLYZfYVRVJWHhJuLmrkTSO1P9Ikb34sz2Iywe9YBL6EEy6p1ZEUWbh6qq0nRYd8b8+QKysWK39IlLtrLzqa/JiU0ioEMUfV659YJsdBFz+DRvP78SFRWHQ0VVYezkDlxxQ28cNjvLJj7Gme1HsOcXovMyIXQSE5a/SWg/1367NpvCfy7/yeMHVKeX+OrXa+roGZ3DfCab7MMn8W3Z1OmIaxB7gRkhS27vp6yDJ1g06gEcZpuzIE0SRE0ZyIifntR0eKrA3p1JzHl9DaoKdpuCwSDTs38Ud/x3KFJZPQ3qEU1aoZqoqsqaa18hYcHG4swZnY+JDrdfQv+3/o+TCzay6c73sGTkoqoqxiA/zGeyUe3nQj2yl4Fuj15Jr2eq1oxEVVXSd8aQF59KSO92+LVsWunrk5dtJ+b7ZeBw0ObacU6tnXqM6dvtDu698Vfy81yzHIxGHfc+PoKuPZuhqiopK3eRun4/3hHBtLpiJIYAX7exVFXlzmvnUZDvnhXRJMyHtz+fUWvPoy7JPHCC9be8SfrOGIQQRE7sx5AvHsQr7NyKUrFYSfhnC4Up6YQP7UZwjzZljKhRHtlZhWxZf4LCfBtdekbQpn2TBr0Xpjn8anJq3V6WT37crQmJ7GVg+s7PCegQhaqqzk1VSfBr9FVu5f5Q9Swb85lslk54hJyjiQhZwmG10+qKkQz58qEKr9o2/t+7HPvfCpcvrJazRzDs60cqbU9NcXBvCu+/ugazh9S1/kNacNfDwys13t+/7WfBr3td8rsNRpnrbuvP8DFtq21vXaCqKpaMHHReRnTerkVI5vRs5re9DmtOQfEehtDJ+LdtxmX7v67TKm6NhouWpVNNEhdtwV7g3nFKVSFp6TYAhBB4R4SUuTlWVg/Tslh73atk7j+OPd+MLacAxWzl+K9rOPzxXxW6PmPPMWJ/WO7yhWXPN3P8l9WkbTtcJZtqApvNUdqeGDZPTUDK4ZKZXZhyWRdMJh16vYy3j57Z1/ZqNM7+1Nq9zO9wA/MiL+fH4OmsvPx5l0yvmG+Xoljtrq0u7Qr5SWdIWb2nPkzWaGRc0Fk6mQdOkH34JIGdogns3LLK4+j9fZD0erdG0pIsueVUG4P8COgUTebeOJfjQi/T4tLK54Bbs/NIWbXLLRNIKbBw6MM/izNUyiJp6TaPdxyK2UbSkm1u8fC6omOXMBQPHbmMRh2Dhpfeb6A0hBBcemUPLpnVjfw8C75+RuQGlhZXGtkxic67yBILi4QFm1ie8iRT1r3vPOfwSZRC9zRVVXGQeywZKlnFm7RsO4c/+xt7XiGtrxxNm2vH1moW0cWCQ3FwKiUXb289gZqWTu1jLzCzYvrTnN50AEkn47ArhA/pypg/X6xStWGbq8ew56UfPD4WfdlQt2PDvnmExaMewGG1o5it6HxMGEP86f3SzVV4LhZEKZtDngqMPKH380bS61DO+9KQDTqPRUBJy7ez69lvyYlNJqhbK3q/eDPhg2tettZo0nPrPYP58oONKIqKojgwmnR07BJOv0Fl69mXhU4nERBYe7o55jPZ5MQk4tsqosYkmA++Px/lvAWFw2ojfVcMmQdOENSlJaH9OnL851VuoUUhRKXj9Nuf+JJDc/4oHuv0xgPEfLuEif++3aBkGBob2zef5JuPNmOzKSiKg7YdQrnroWH41+L7sTLUyPJHCDFRCHFECBErhHjMw+M3CiHShBC7i/7dWhPzlsbWhz9zluMXWJwhkAILp9btY/tjX1RpPN8W4Qz75lFkbyN6P2/0/t7o/bwZ88cLGAPdNxBDerVj5tHv6fns9bS9cQL93/4/LjvwtcvGWkXxahqMKdyDU5EloqZ57o16Pi1nlRILF9DqipEuh078vo5/L3uGtM2HsJzJ5tSq3Swd9zCn1pQfMog9nMa7L6/isbv+4ssPNpKaklPuNQOGtuSl96cyZWYXxkzuwD2PjuD+J0c1uIIVcIrZbbzzPX6JvpJlkx/nt9bXsPrql9wcdVXIPpLgssl/FkknkxfvVDZtfc0YDIG+zraIRcgmAyF929OkX4cKz5WXcJqD7/7mFuJL3xnDyb82VONZXFxkZRSQm3PuNTx5PIPP3llPXq4Fi9mO3eYg5tBp3nphZT1a6Uq1V/hCCBn4CBgHJALbhBALVFU9eN6p81RVvbu681WE2O+WusXLHWYrsd8uYeD7VTOh1eUjaT6pPymrdiN0MhGje6Ero2rSKyyI7o9eVaW5SiKEIHRAJ/LjU92OV3R8U2ggo399llVXvFh8t6AqDkb8+CTeEeeEoFRVZesDH7tVtyqFFrY9/BlTt35c6hw7tybwyVvrsNoUUCE1JZdtm+J55o1JREYFlmlfeIQfM6/uWaHnUp/sf+sXYr9fhmK2Fr+/Tv61ge2PfcGAd+6s1tjhQ7qSun4/Dsv5q3w7Qd2c4S29jxdTt33C9kc/5+SCjcgGPW1vnECv526oVNbIqdV7EHodnDeXPd9Mwt+baDmzcpvlFxvHY9P57N31pJ3OAxVatAnm/x4YxrK/D2M7T51WUVRSkrJJOJFJVMv6z82viZBOfyBWVdU4ACHEz8B04HyHXyeoqlrq5qi9lP6vFUXv5030tMHVGqOy5CemkeBh1SUZdCSv2IEh0I+CpDOEDuhI6MDOpX7wm08awFWp80lZuQtU1fmFdV4WiGKxUZB4xuP1GfviPB4H52v+/adbsFpLNlpRMZvt/PL9Lv775KiKPNUGz8H3f/fwZWjlyOcL6f/WHdXKkul016Uc+ugvrHYFVXE6DdnbSKvLR+IbFVZ8nnfTYIZ/53YTXSkMgT4e3ydCJ2MMDUBVVfJPngacd7ca58jNMfPa08tdMsuOx6Tz8hNLCWvqh+pwz3qUZYmszMILxuFHAgklfk8EBng4b6YQYjhwFPivqqoJ558ghLgNuA0gOrpqMVwhBOHDupG6dp9rab4QNB3RvUpj1iepG/YjGfQo563GlAILm+/9ENmgQ7HYkfQyYYO6MHbhy6VmCum8jERNGVjqXLJRj87PC1t2vttjJe8Ezicv10JujvtmIirEHDpd6nWNDWuWZ20kxWzFYVeQDVV3+KbQQKZt/5Q1173C6Y0HnGqfkkRIn3aoqlqjed+RE/ohGdw/+pJeR/iQrvze+SbyTzrvKH1bhDPy56cJ7q7l8gOsXxWHoriu4h0OlcICKyGh3ugNsltzF7tNoUXrhqFOW1eB0r+BlqqqdgeWA995OklV1c9VVe2rqmrf0NCqVyYO/ug+9P7eSEan45NMBvT+3gycc2+Vx6wvTKGBeFRwwqn1bsstxGG1Yc83k7phPwff/73Kcwkh6PrQ5W5yBjpvIz2eKr1S1WjSe1KXAMDPv/E1tCiN0IGdPB4P7NSiRrRqMvYcI31nTLG0s5JXyI5Hv+TgnKr/TT0hG/RMWPYGXk2Di/ekdN4mBn5wN+tvfpOcIwkohVaUQivZhxNYPOoBbHkVSxC40Dl9Ktdjty6HohLVMggfXwM63Tm3ajDKjLukY4Np7FITDj8JiCrxe/OiY8WoqpququrZJeCXQJ8amLdUAju3ZMahb+n28BU0v2Qg3R+5ghmHviWwY9UzP6pCwakMDn/2N4c+/ov8xLRKX6+qKum7Y7Dlu9cAeEIptHD068Xlnpex9xhbH/6UTXe9T/K/O12EoHo8fjVd7p+FzseE7GVE7+9Nr+dvpN1NpXfuMRhkBo9ojd7gmt1hMMpMmXHhNKUe8O5dTgnmog1lIUnovE0M+ui+Ghl/x5NfuYWM7AVmdr/wg4saa00Q0qsdlyf8zLhFrzLq12e56vR8hCR5TN91WO2c+G1Njc7fWGnfMQyjyUNgRECnrk158Z0pjJncgbCmvrRqG8LNdw3i8usbjiRKTYR0tgHthBCtcDr6K4GrS54ghIhQVfWs/OM0oNY7aHs3Dab3CzfV9jSlEvP9Ujbd8Z5TT0eFbQ99Sp/X/0OXCuTNn2Xfm/PY88IPLs08ACQvA9gdHj+cnjI9SnLgg/nsePwrHFYbquIg9vtlRE8fwvAfHkcIgZAk+rx0Mz2fvhZLeg6m0MAK5WZfe1t/zGY7O7acRKeTcSgOJk3vzLAxF04oILhHG6bv/Ix9b/zMmW1HCOzSku6PXUVQ18rXDHgi7/gpj8dtOQXYCyzofWsmtS/3eAr5CWkEdW1J+JCuxcfzk864KHyexV5ooSApvUbmbuz0HRzNX7/sJS01r7iLnN4g075TGK3aOsOeV9/cl6tv9ljoWu9U2+GrqmoXQtwNLAVk4GtVVQ8IIV4AtququgC4VwgxDbADGcCN1Z23IWDLK+TEb2soSE4nbFBnmo7siRCCgpR0Nt3xntvm8Y5Hv6D5hH4EtI8qZcRzOBSFva/+5LHCN6B9c5QCCzkxLjdSSCYDba4ZU+qYBacy2PHYly522fPNnPxrAyn/7qTZ2D7Fc+fEJqPzNla4EMdgkLnzoWHkZJvJTC8gLMIPL68LT5LXv20kQz5/sFbG9mvbzK1gD5ybrDqf6ocErDn5rJr1PKnr9yEZ9TgsNjrefSn9Xr8NIQRhAzuh8zFhPy98o/M2lRrOuhiwWhXMBVZ8/U3o9TJPvz6Jv3/bx5b1J9DpJEaMa8fEaY3j9amRwitVVRcBi8479kyJnx8HHq+JuWqb1A372fHEl2QdjHc2IX/hRppP7O92XsaeY87iKpu9uP1ek74dGL/kNeL/WO9RMtlhVzjx21p6PFG+cqMtp8BjVSU4V4IT/32bJWMexGE71+3Jv20zupbR1Sh52XaXHO6z2PPNnJi/lmZj+5C0fDtrr3sVe76zqYt/u+aMnv88/m2alWszgH+AqcHEKxsbfV+9lZWznnf5u+u8jfR+/qYa2bRdf+tbnFq3F4fFVvylf+STBQR2jKb9zZOIGN2L4B5tSN8ZU2yD7GUkuGcbIuqgF29Dw2ZT+PGLbaxfFYeKio+PgWtu7ceAoS258sY+XHljrUama4WGV91Sj5xas4el4x8hdd0+LOk5nNl+hJWznuP4efFLVVVZdfnzWLPynI5RcWDPN5O29TAH5/zhTKvzJEqnqh7DMJ4wBPig8/F8Cx/QIYomfdozO+5H+r7+n+JuT1O3foK+lGvAKdvsUbxGktB5G8k9cYqVlz2L+XQW9nwzSqGVzP3HWTLqARxK5bVtNCpH80kDGPnzU/h3iELIEj5RYQyYcw8d/29atce25RaQsGCTW56/Pd/M/nd+BZx7EhOWv0nPZ67Dv0MUAR2i6Pns9UxY9maDVoesLb79eDMbVsdhsynYbQ6ys8x8OWcjh/enln9xA+WClFaoKlsf+tRtVa0UWNj24Ke0nDm8+E2fdzyFfA/56kqhhdhvlzJ24Stsf/Rzt8clg54WHqQYPCEkiV7P38COx75w7ZzkZaTPK85CZWOwP53vvqzCzy9q8gC3/QBwpmO2uW48R79ahMN+3heSQ8WanU/Kyl1EjmuYcckLieipg4meWvO1Htac/FIlOqzp5yqidSYD3R+9qkaKBhsz+XkWtqyPdxPxs1oU/vp1Lx27jqsny6qHtsIvQdaBEx6PFySfQbHYyIlN4sB784n9YXmp7fdUFfxaNqXXCzciexkQOtnZqMLbSOf7ZlQqn7nz3ZcxYM49+LQIR9LrCOrWijG/P1/lRih6P29Gz38enY8JvZ8XOl8TsklP7xdvIqRnW/LjU3FYPWwEOxxujbE1Gi65J05x+NO/if1hebHapndECMZgP7dzhSzRbFzjC03UNlmZhcg6z1+Qaacq36u6oaCt8EvgFRHsMVNC5+vFvjd+Zt9rc50pjELgsLhX7cpeRtrdNAGAbg9dQdSUgRz/ZTWq3UGLmcMI6Vl5md72N02ifRkpkZUlcnxfrkz+lYR/NqOYrURO6FdcVNVsbB/i/1jvJs6lKg7CakE8TaNy2PILcdgUj/pNZ9n13Lfse2MeCIGQJTb933uM/v15Isf3ZcCce1hz7SvOsI5DRTLo0PmY6FWP2WwNldAwX1QPmbBCErTp0KTuDaohtAYoJYj5dgmb757jkhkjextpd9NEYr5e4nETVfY2OjdNfbwI6dWW8UvfKFNjpyFjN1v5u+8d5MalFG/qNYRGKRc75jPZrLvpdZKX7QDAv0Nzhn39CE36ugqmnd50gCXjHnbL5Ze9jYT0bEvaFmc2tCksCK/QQCLG9abrf2fh3azxOrDaZOH8/fz1S4mGOsIp3f3cW5Np1jygfo0rg7IaoGgr/BK0vWEC1qw8dj3/PQ6LDaGT6PLfWU6xLIu7IuJZrRO/VhGEDepMxJjejXpzS2cycMmmDznw3nyOz1uFzttExzun0fb68fVt2kWLqqosGfMgWYdPFvdEyNp/giVjHmTGoW9dnHXMd0tRPOhFKQUWTm86WByGNKdlIRv19H31P5oUchlMmdGF4Cbe/P3bfnKyCmnbIZTZ1/Vq0M6+PDSHXwIhBF3un0Wnuy/DfCYbY7AfskHP1kc+8xizF0IQPqQr7W+ZXA/W1g56P296Pn0dPZ++rr5NqVPiYs7w+097SDiRSXgzPy67sgedulWub3BtcHrTQXKPn3JvgGO1c+SLf+j17LkeyQ6LvdS9pfO7ZFnSs0laspWoSyomsX0xIoRg8IjWDB7Rur5NqTG0TVsPSDoZ76bBxfoorWaPQPZyD9OoioOoS0oXI6tpCpLPkLJ6N/lJlZdpaGwUFtqwWiqWwlpdjh46zatPLWPfrmSyMgs5cuA077y4kh2bT9bJ/GWRF5fs8bjDYiP7sKt9ra4YWeECLcVqJyfW89gaFy7aCr8ChPbrSKc7p3Poo79wWG0ISULoJAa8dxdenpqT1DAOm511N71B/Py1SCYDDouN6OmDGfbdY6WKduUnpmHJzCWwY3SjaluXGJ/Jl3M2EX88AwF06RHBLfcMJjCo9joGzf1mh0vjc3BWV/709XZ6D4iq1zBdcK92xXLJJZG9jW4b6ZET+hE9fQjxf653i+Ofj6TXEdT9wlm5alSMxuMJKkjGnmOkbTmEd7MQIif2r7EYZb83bqfNtWM5+ddGJIOuOHZfF+x67jvi/1iPYrEV7yWcXLCJnU99Q783bnM5tzA1g5Wznid9x1GEXkaSZQZ9cj+tr2j4mvR5ORZefmIpBfnn9kv270nhlSeW8tpH05FKySOvLgknMj0eT08rwGZzYDDUX5w7qEtLmo3rTdKyHTjOSmLIEoYAX5pfMohtj35O4qItmMIC6frfWQz/4XF2PPkV+9/6pVRdJcmoJ6BDFBGjetbdE2lAqKpKUkI2FrONFq2C0ekvnn2MC8bhO+wKq654gaQl20CAJMvo/byYvPY9/FpXTBagPIK7t6kXXfDDnyxwLwgrtHDks7/dHP7yKU+QsTfO+WEvSjZaf8ub+Ldp5pbVUZ+oqsqyhYdZ/McB8vKstGobQss2wdht52mNKyrZWYUc3JtC154183c8H/8AE+lp7j0AjCadi9RtfdH5nhnODB0hoEgbv+vDl7N42H2Yz+TgsNrgAJzZcpgeT19L/Py1pTp7Q5Afba4dS5+Xb2nUCQZVJTUlh3dfWkXGmXznnbqAm+8eRP/BLerbtDrhgnH4hz9dQNLSbcWOUcGZt7zy8heYvv3T+jWumpSmRW7LK3RpjpG5/zjZh917oypmKwfen8+IH56odVsryi/f7WTF4iPFoZSjB08TeyQNh+K+6ehQVNJSa6/YZeqsrvz09XaXsI7BKDNhascq3VVYMnNJXb8Pg78PYUO7IslVX0Fac/L597JnXATvVLvC9kc+R8iS09kXYS8ws/v576GUfsCSXmb28R8x+PtU2Z7GjENx8NrTy8lMLyjaw3b+vb94bwPNowJpFlV29o3VqrB53XH2704hpIkPI8e3IzzCvZitIXPBOPwjn/7tHrd0qGQfjCc/MQ2f5lVvqFLfhPbv6OyCdB5N+nVwWaUVnspAeLo9dZxrWVfXHD14mt/n7iE5MZvm0QHMuKonkdEBLF90xK2RhOpQkSSB4zz5ByEE0a1qrz3cyPHtyMu1sPC3/aiq8+5j9MT2XHpF5TukHZjzOzse/cLZUUp11jGMX/o6wd2qFi8/+ddGj8dVRfHc9Nygx7dluEfVTVNoIHo/7yrZcSFw5OBpCvKtbolMdruDlUuPcu2t/QCwmG2sXh7Lzi0J+AeYGDu5A9Gtg3nh4cWkp+VjsdiRZcGKRYe559ERtGgdzKF9pzB56enaI6JBh4guGIfvKU8eAEmU/lgjYeCce1g04r8oZiuqXUHoZGSjnkEfunbwCundzk0cC0A2GYic0K+uzC1m365kPnhtdfHKOTuzkJhDy7nm1r7IsoSN8xy+s4gZWRYoRSt9vUGmVdsQWrerveIgIQRTZ3Vj4vTOZGUUEhBowmCs/Ecjbcshdjz+pUuTc1tuAcsmPMrlCT9XaaVvzc73HJ4pJfvSYbPT48lrWXfT686c/CLvJnsb6f/unRdlGOcsOdmeGwk5HCpZ6QUAmAttPP/wYs6cznP2aBawe3sinbo2JS01r1hbR1FUFEVhzutrcDhUZy9tu/O1bt0+hNvuG0JEZMPL16//AGUN0frq0chG94wVr7Ag/FrXzeZqbRHSqx3Td31O+1sn06R/R9rdPInpuz53i8kbg/3p+vDlLql5kkGPMcS/RhQXK8uPX27zmP2ybMEh7J6cmIAuPZoyZGRrfHwN+AeamDC1Iw8+M7pOHJVeLxMa7lslZw9w+NO/PRY+2fPNpK7bV6UxI8f18ahwKpsMHu/m9P7etJw1nMlr36P5xH54RQQTNqQLY/54gVazR1bJhguFdp3CUOzuGU8Go47ufSMBWL089pyzB1Cdgml7dya5CamB8zG7zVHs7AHijqbz3EOLyMooqJ0nUg0umBV+t4euIP739eTFn8KeZ3Z+IGSJ4f974oJY1fi3acbgj+8v97xez91IcI+2HHj3Nyzp2URdMohuj1yJMahuY42qqpKSlOPxsaTEHIaPbcvmdcdd4+YGmRlX96R1uyZcdlU+CfFZhDWtugOuayyZOZ4Ln4Szv0FVCOgQRftbpxDz9eJijSOdj4mIsb1JWbkLu811f8eWW0Dioi1ETRnIuH9erdKcFyrBId6MmdyBVUtisBTVeOgNMmHhvgwc5uxatnPLyXPOviSCUu+qPGG1KqxYfJRZ1/SsvuE1SOP4JFUAvZ8303Z8Svz8dZxauwfflk1pe8MEvJs2jG7xdYUQgpYzhtFyxrB6t8PXz0herud88PAIP8ZO7sC/i49itdgJj/Djutv607J1MF+8v4Et6+PR6SUUu4M27Ztw35OjGnwHrRaXDSPl311u4nMOq53wYd2qPO6A9+6i+aT+xHyzBIfVTptrxmAKC+LUyt1u5yoFFo5+uYioKXVXENiYuPLGPrTrFMa/i45QWGBjwJAWjJ7Uvjj11s/fc+GaLEsg45ZFVhoOReV4jLuEen1zwTh8ANmgp/VVo2l91ej6NkUDmHhpJ377YbfHx5b/c5j3v57F7Ot6oygO9EXhiX/+OMDWjU4d8rO30DFH0vjuky3c8UDFegmcJS7mDJvWHgcVBgxtSduOtbtx3/qq0Rz5bCGZ++KcTl8IZC8DfV6+pVp3WEIImk/s79J5LfnfnZ6b2eDsQavhGSEEfQdG03dgtMfHx13Skb07k1zuPIWAkFAfOnYJZ+Pq48iyAAE6nYzFbPcY6pFlQXSrhrfYvKAcvkbdc3h/KvO+20HiySyCQry57MruDBruzEgZP6V0h1+Q54x1S5JAks7Folf8c9gt7m+3Odi2MZ5b7hlU/MVQHvN/3MWSBYeKM4FWL49h1IT2tdpcWjbombT6HeLmriR+/jqMwX50uP0SwgbVvLR02OAuqB6a2eh8TLS5uvS+xhpl07FLOLOv7cUvP+xCp5NwOFQCg714+NkxhIb7MXVWV2IOpxEQ6EX7TqG8+/IqDu5Ldftb6A0yYyc3nLqXs2gOX6PKHDmQytsv/Fsc80xNzuXrjzZTkG9nzKT2GE06IiL9Pcby23bwvNouLPCcUaWqKjarUiGHfyoph8V/HXJJ+7RaFFYtOcrQUa1rdeUlG/S0u2EC7W6YUGtzAOi8jAz75hHWXv8aql3BYbOj8zURNrALrc9z+M5eyms4/stq9H7etL91Mk2HdceSlUfsd0vJ2HOMkF5taXv9eAwBpWvtXyyMn9qJYWPacOzoGXx8jbRsE1y8Dxga7kdo+Lm7tYeeGcOWDfH8Pnc3aafyUFVo074JN9wxgJDQhlfvoDl8jSrzyw+73Da4rBaF33/azagJ7ZAkwQ13DOCdl1ZisyqoqnNFrzfIXHWz5y5LXXpGsGNzgtuKKaypH94+FeszsGtbosfVr92usHNrQoO81a4KLWcOJ6R3O2K/W4o5LZvmUwbSfGI/hHQu+c6hKCyb/Bhpmw4Wh5ni56+l/W2XcOyH5dgLLSgFFk78sprdL/6PqVs/xq9l/auE1jde3oYKVXZLssSg4a0YNLwVqqo6a0lKKXxrCDRcyzQaPEknszweN5ttxSGbTt2a8vRrE+k/pCWh4b74+hkRAj57bwPbPahRXnF9b7y99ej0zremJAsMRpmb7qz4JqTeIHvs3ypJEgbDhbXG8WsVQa/nbmTQR/cRNXmAi7MHSFiw6ZyzB1BV7AUWDr7/O5bM3OJiRXuBBWtGLpvv+aCun8IFgxCiQTt70Fb4GtWgSZgPCSey3I7rdDJePucyaqJbBTN2SgfefDah+I4gMT6Lz95dT+Ft/Rk25lzrx7Cmfrz64TSWLzxM7JE0mjUPYPzUTjRt5l9hu/oOiubnb3e4HReSoP8Qd80Um01h5ZKjrF95DCFg+Ni2jBzfvkHo6FSXk39tcMsaApzpo+fdBKkOR3FXrYZMfFwGq5fFkJttpvfAKPoPbtGgq1sbEprD16gyM67uySdvrXMJ6xiMMpOmd3KmsZXgl+92egz/zPtuJ0NGtXHRrAkI9GLWtb2qbFdgkBe33jOYL+dsLB7X4VC54fb+NAlzjVE7HCpvPf8vcTFnijeL5323k93bkuqs4Ks20Qf6ImTJo8SyJ6Q6vgNKT8tnxaIjJMRn0rpdCGMmdSAgsHQp7LXLY/jhi23Y7Q4cDpW9u5JZsegIj780vsIb+hczmsMvg+wjCeQnnSG4e2tMTRpemXR907t/FDfdNZB53+4kJ8eM0ahj8mVduGRmV7dzE+KzPI5RWGCjsMCGj2/N9gEeOKwlXXtGsGd7EioqPfpEesyxPrg3heOx6S6ZQVaLwtGDp4k5nEb7TmE1aldd0/6WSRz94h83tdWzjt1hPddkRjLq3TZ8a5MTx9J55cll2O0OFLuDw/tOsXzhEZ57axI6ncyZ03lERgfi62cEnE1xfvhim8vCwWK2k3Aik01rjjN8bNvSpqp3MtILiDl0Gv8AEx06h9Vb6OeicPhJy7ez8+mvyYlJJqBjFH1eupmIUaWvIC0ZOayY/jTpO2OQDDocFhud7rmMvq/9p9Gv+GqawSNaM2h4K6wWO3qDrlR1yZBQb5JOZrsd1+kkTF618zb09TMyZFRr9uxI4tWnlpOanENwE29mXN2jOHX06KHTWMzunbVsdoWYQ6cbvcMP7taaAe/fxZb7PnQ2wlGdzn7kvGfY/uhnZB9JLK4ODuzckv5v3VFntn3z8WaX195mc2C3W3nx0SWYzXZ0Ogm7zcGYSe258qY+xB5OQ9ZJ4OFOccv6Ew3S4auqys/f7uDfRUectgPePgYee3Ec4REVD1PWFBe8w09YuIlVV7xYvMJJ23SQ5Zc8yZjfny9VUGzNda9yZuthHDZ78XWHP/6LoK6taHvduDqzvbEghMBoKrsK9rIre/D5+xvcJYinuYd/KkJ8XAbJCdk0iwqgRevSs2727kziw9fXFK8KT5/K4+uPNmOzKAwf147AIG8MRtkt91+vl8sMLTQmOtw6hVazR3BqzV503kaajuiBpNcxdesnnN54gOzDJwns3ILQgZ3rbEFjtynEx7k3nlFVyM1xfubOptWuXHqUppH+REYHopbSs7eiGVx1zY4tCaxaEoPN5sBWVKVrMdt596VVvPrhtDpfQF7wDn/rg594bB6y9aFPuMyDwzenZ5OychcOm+uqz55v5sC7v2oOv4r0G9yC/Hwrv36/C3OhDVknMWFqRy69sgcAx2PT+eX7nZw4lk5QiDfTL+/OgKEt3caxmG2889Iq4mLOIAmBQ1Vp2SaEB58ejcmD9MKvpaSO/vrjboaNbcuAoS2Y991OOE+5U5YEfQd5rsZsjBgCfImeNtjlmBCC8CFdCR/iHoKrbSRJIMsCu718gRqrRWHJX4d49cNpeHnpMRe6fjYNRpnRE9vXlqnV4t9FR4p1e86iqpB+Jp/kxGwiowLr1J4L2uGrqkpOTJLHx7IPJ3g8bsspQJSy4rSk59aYbRcjI8e1Y/iYthTkWzF56YuzYJyx3KXFq+yC/Gy+nLOR3GwzY6d0dBnj5293cuxIWvFqCZwSCnO/2eExdfNUKQJueTkWrBY7Pr5GHnl+LB++sYb8XCsqKgGBXtzz6AiPXyAaNYMkSwwc1pLN609USJ8mP8+CJAkefGYMbzy7AmvR3oNid3DJzK506tYwawdKKySUJMnti6suuKAdvhACU2gA5jT32LFXuOeGGr4twtH7ebs1UxE6mchJ/T1eo1FxJEkUb8KdZf6Pu91llC0K83/azcgJrumRG1bHuTh7cEovbFwd59Hhh4T5kJLo7vS9vPXFKpxt2jfhnS9mOM8TEBHpr+3V1AHX3dafM2n5xMWcQZad8XqEis3q+vcVkih26FEtg3jv65kc3p9Kfp6VDl3CGnTorf+QFiSezHJr9iMEtKjFpj6l0fgTjcuh++PXoPN2zc7QeZvo8dS1Hs8XksSQzx5A9jYWF7FIRj3GID96Pu35Go3qceJYhsfjdpuDnKzC84557tVqsyke47uzrunl1oTcYJSZfnk3F6cuhKBZVADNmgdozr6OMHnpefyl8Tz75mT+c+9gXv7gEu55dCQGo8zZP4EsS3h56VzSdGVZokuPCPoPadGgnT3AmEntCW/qV7y4kCSBwSBz892D6qV24IJe4QN0vm8G9gIz+17/GYfNjmTU0+PJa+hw+9RSr4meNpgp6z/gwLu/kRuXTMToXnS++zJMoYFlzpUbl8zeN+ZxZuthAru0oNsjV1a5tV1FsZut7HzyK45+vRilwELTUT0Z+MHdBLSPqtV5a5ImYT6ldiM6/26gU7emHNiT4iI7L4TzuCdH3XdQNFbrQH75fhdZGQX4+BqZfnk3xl3S0e1cjfqheXQgzaMDAQiP8OepVyey6I8DpKbk0q5TKJOmdya4ScPTpakIRpOeZ9+azOZ1x9m7PYnAEG9GT2hfZv/c/DwroOLjayz1nKoiStv1rm/69u2rbt++vcbGc9jsWDJzMQb7I+lq/ps1Y18ci4bei72wqA2hJCGb9Ixd+AoRI3vW+HxnWTblcU6t2n2uybUQGAJ8mHHoG7zCG4dmzJ4dSXz4xhq3DJ6R49pxza2uG+unknN44ZHFWK2KU0zNIKPXSzzzxqRyW8rZbQqyTrqgVvB2sxVrVh6m0IBqNUvXqDjZWYUs/G0/u7cn4etnYMK0TgwY2rLa76u01Fw+e3cDcbHpAES1COT2/w6lWfPK1QAJIXaoqupRFvaicfi1zZLxD5OyYqfb8YCO0cw4+E2tzJl1KJ4Fff/PvajGZKD7Y1fS65kbamXe2mDDqmP8/O1OCgqsSJJg9IT2XH5Db48pm7k5ZlYviyH+WAbRrYMZOb4d/gGeG1d44sSxdP5dfJScrEJ69Y9i8IhWjaar1lkcNjtbH/yEo18tKm6W3u+tO2pUpdOWX8ju578n9vtlqIqDFjOH0+flmzGFXLxFiHm5Fp68929ycy3F7RKNRh3jLunA7Ot6F5+XkV5AeloeEZEBbnepnrDZFB667Q+ys83nhP8E+PgYePuLGZVq/lOWw29c7/IGzOmNBzwez4lJxF5oQedV87dnWQfjkfQyimuYG4fZypltR2p8vtpkyKg2DBrRmrxcS5F4WumrVT9/E1NnVa2D1NnSfJvdgepQObjvFMv/Ocwzb0zC2Iic/ub7PiT2+2XFPXQVs5VNd72PKTSQqMkDqj2+qqosHfcwGbuPFd89xn6zhJR/d3LZ/q+QjQ0z7722+XfxEfLzLC69cS0WO0sXHGbi9M4YjTo+fWc9e3cmo9M7N6JHTWjHVTf3LbUoEWD3tkTMZpuryqsKdruDLetPMHJcuxqxv0Y2bYUQE4UQR4QQsUKIxzw8bhRCzCt6fIsQomVNzNuQMAR61hGXDLpa0ycJ6BCFw0MzcMmoJ7hnzbxB6hJJEvgHmGplM8tuU/j+sy189dFmrFal+INltSicTsllzbKYGp+ztrDlFxL77VK3TDKlwMKeF3+okTlS1+0jc//xc6FCnHcVhamZnJi/rkbmaIzs353iliUGoNNLxMdl8MMX24obnhcW2LDZFFYvj+HfRYfLHDctNc8tkwecRVppp2ouHbzaDl8IIQMfAZOAzsBVQojO5512C5Cpqmpb4F3g9erO29Doct9MZG/XVbzsZaDdTZNqLbYa1LUVYQM7Ixldb/dko55O/1f6pnRDQ1VVCvKtpWbgVIb8PCupKbko54mFffPxZtYsj/V4jdWqsGntcX79YSdvPLuCX77fScaZ/GrbUtPkHEtm+xNfsv7mN1EdnnPX806m1shc6bticHj4e9jzCknfcbRG5miMhIb54ClUrygO/PyNbFrjnjZstSgsWXCozHFbtgn2uNAxmXS0atukWjaXpCaWnv2BWFVV4wCEED8D04GDJc6ZDjxX9PNvwIdCCKE21A2EKtD1wdnknThFzDdLkIx6HBYbUVMH0f/t2tUmGfPXi2x98BOOfb8cxWojfHAXBn50H97Nau5NUpvs25XMd59uIeNMPpIkGDq6DVff0s8tlbI8LGYbX364iZ1bEpBlCZ1O4uqb+jB0TFvycixsWX8Cu730Ap8TxzI4eTwTu93BkQOp/Lv4CE++OpHolpXPld67M4nlCw+Tl2uhz8BoxkzuUO0G7PF/rmfNta/gsCmotlIKdoSgSd+aaavn1zqiWEeqJDofE/7tmtfIHPVNbo6ZX77fxfZN8QghGDyyNTOv6Vnm32rCtM5s23TSJcFAlgWR0YGEhvvh8NB4B85m3pROp25NaR4dyMnjGcVfGDqdRHCoDz371dzrXe1NWyHELGCiqqq3Fv1+HTBAVdW7S5yzv+icxKLfjxWdc+a8sW4DbgOIjo7uEx8fXy3b6gPzmWyyjyTg16ppnTpdVVVBVd0aYDRkjse6VtiCs3lJr37Nuevh4ZUa64PXVrN3R5LL6spglLn/iVH4+hl59allpVY9CgGePgbtO4fx5CuV2wRd8Mte/p6/v/g56Q0yTUJ9eP7tyeXqDZWGYrEyN3wmtpyC0k8SAp23kSkbPiC4e5sqzVMSh13ht3bXUZCYdk5aWQgMQb7MjvsRg3/jTJM8i82m8MQ9C0hPKyi+G9TpJSKjAnn+7cllZtxs2xjPt59swWZTUBQH7TqEcufDw/HzN/LwHX+Slprncr4Q0KNPJP99anSZNlksdhb8so/1q46hqk7F10uv6F5pnaBGs2mrqurnwOfgzNKpZ3OqhKlJQL1IKQsh8Hiv2YBZOH+/W9zSZlXYtTWBrMxCAoMqVlSTk21mz44ktxJ9q0Vh4e/7uffREaWu7oVUircHYg6dRlXVCqfb5eVYWPDrfmwlQiE2q0L6mXzWrDjG+Crm/p/ZXnoIRfYxoTMZaNKvA31eubVGnD2ApJOZsv591t/0JqfW7EYFmvRuz9BvHmn0zh5gx+aTZGeZXUJ/dpuD1OQcDu07RefuEaVe229wC3oPiCI1JRdvH4PL+/TG/xvA+6+uPtfSUxYYDDquuMFzS8+SGI06Zl/Xi9nXVb0XRHnUhMNPAkpW+TQvOubpnEQhhA4IANJrYG6NRsyppByPvlanl8k4k+/yQbJY7OTlWAgI8nLrRJWTVVgspXs+6afz8fI2MG5KB1YsOuJyN6HTSzz6wjjefmEl5kL31b/BoKtUbvWxo2fQ6SUXhw/OL57d2xOr7PBlk8Fjj16ApkO7Mn5x7WyJ+USGMmHZG9gLzKiKA72fd63MUx+ciM3wKItttzs4eSKzTIcPzmpfT/nxXXs246lXJ/LP7wc4lZxNm/ahTL6sC6HhDaM5fE04/G1AOyFEK5yO/Urg6vPOWQDcAGwCZgErL6T4vUbVaNOhCcmJ2W5xT7vdUawVbrc7+Omrbaz919l+UKeTmHlNL8ZOPherDmvq5/GLQ5IEHTo79ewvv743QSHeLP7zIHm5Ftq0b8JVN/WlRetgRoxry8rFR10ctV4vMWxM5VbLvv5GjzFcIajw3YonQnq1xRjkiz3PNf9W52Mqs2K8pjhfmuRCoGlzf4xGnZuSpU4vEdbUr1pjt2gdzJ0PDavWGLVFtQO+qqragbuBpcAh4BdVVQ8IIV4QQkwrOu0rIEQIEQs8ALilbmpcfEyZ0RWDUYYSi2iDUWbMpPbFHbDmfrODdf8ew2ZVsFoUCvJtzPtuB9s2xpe4RseMq3s4xypCCDCadEy7vFvR74Lxl3Ti3S9n8sW8q3nsxfHFOvqzru1Flx5N0RtkvLz16A0ynbo15YobzhXSVITW7UIIDPJya6CuN8iMm1L+6j4/z0pSQhbW85yQkCTG/v0yxiYB6P280fmYkE0G2t08iejpQyplY0XISzhNwsJNZB44UeNjNxQGDm2JvoRmDzgXCD6+Rnr0iaw/w2oZrdJWo15JPJnFvO92cvTgaXz9jEy6tDNjJrVHCIHVqnDntfM85idHtQzkpfdcV7c7Np9k4fz9ZGUU0rFrOJdd1aNSq7VTyTkkJ2bTLDKAppFV60aUlprHOy+t5MzpPGRZwuFQue4//VwatZ+Pzabw7ceb2bz+BDqdhOqAqbO6csmsri4hJcVqI2npNixncmg6ojt+rZtVycbScCgKG/7zNsd/XuXMNLPZCenVjnELX8YQ0DBCEtXBblPYtS2R06dyiW4VTJMwH775aDMxh9NAQJfuEdxyzyCCght36EqTVtBolGRlFvLQbX+4xcTBKar20Q+X14NV5aOqKkkJ2RQWWGnRKrhc2YZvX1vGwb+3U6AzkRvYxJkNY5S54Y4BDB1VM5uwFWH/u7+x8+mvXQq6JIOeqKkDGf3rc6Vel5FewLEjafgFmGjfKazMitL6Ij0tnxcfW0JhgRWrVcGglwmL8OOJl8cjyxJCEhVqgu5QHKxYdIR/lxzFarbTZ1A00y/v5rFfcn3RaLJ0NDRK4u9vxGjSeXT4rdqG1INFFUMIUaz+WBaqqrLp3g9RPllAO0kGVcXi5cOeQeOx4sPC+fvr1OEfmvOHW/Wuw2oj4e9N2AvMbrF8T/1afX2NPPriuGrHwWMOn2bV0hjyc630GxLNwKEtq1WB/eWcjWRnFhbvsZgVO8mJ2cz/aQ/X3uq51WlJVFXl2JEzfP/5FpJOZhdnfa1acpSdWxJ45YOpjaJhTuNJ2ta46JBkiatu7O0Sm0c4Y/a1mbpWV8T99C+x3yxBcjjQ2W3oFDteeTl02b4agJwsz5LRtYUtt/Q8f/t5An0AOzaf69dqLrRjLrSTfiaf915ZVS07Fv91kDeeXcHG1XHs3p7I959u5bWnl5dZOFcWVoudIwdS3ZMDbA42rTle7vU2m8Ibz67gtWeWEx+X6WKH3e4gN8fMhtVxVbKtrtEcvkaDZuiYttz18HBatwvBP9BEzz6RPP3ahDIblzcWDr7/O0qBq1OXUPHNzsBozqdNh7qtlo6c0Ndje0/f6HCMwe57Gsv/OeyxX2taah4pSe5d5ipCXo6F+f/bhdWiFGdeWSx2Th7PZOuGE1UaU6XUUotSm6KXZPGfB4k9nOZxLwmcabdHDtSMpEVto4V0NBo8Pfs2p2dfZ3l5RnoBKYnZpKXmEhpevbBBfWPN8azXowqBl3Bw+XWVyxKqLn1euZWkpdux55tRzFaETkY26hny5UMe6xHK6tda2mNncSgOlv9zmBWLjmAutNOzbyQzrunprGXQyW56NBaLnW0bTzJ4ROUbChmNOtp2DC0qpDt3XJYF/Ye0KPf6tStisZbi7MGZKlzdEFZdoTl8jUaBojj4cs5Gtm2IR6eXsdsddOnelLseHt7otOzP0uKyoRx4b76bXo1sMvDIp1cS2aJue576Rocz4+A3HPpkAanr9hHYKZpO91xGQCnaOf0GR5OcmO2xX2t0q7LvwL6cs4ltm+KLC+HWr45j944kbrxjAJ7W3EK4dz+rDLfeM5gXH1uM1aJgMdsxeekIDPKuUGiwNH2cs8iyxKgJ7atsW13SOD8pGhcdC+fvZ/vGk9hsjuLV34E9p/jp6+3c+H/uzcsbA90euZLj81ZTmJqJUmhByBKSUc+ouU8Q2aLmQ1bJidn8+OU2Du9PxWjSMWJcW2Zc3dMlO8UUGkivZ66v0HjjpnRk4+rjpKflYbEoSJJAp5O45e5BHNp3ip+/3cGppByCQryZcXWP4tV5WmoeWzfEu2zGOxSVwgIbyYnZGAyyW+Wz3iAzakLVJb/DI/x4+7PL2LrhZFFaZhC9+ke5VW17YuCwliz9+5DHSu7QcF9uu28IIaGNQ25CS8vUaBTcc8OvHvve6vUyn8+7qkGmAlYEW24BMd8sIWnZdnxbhtPprksJ7FR+mKGyZGUU8NjdCzAX2orDGnqDTNceEdz/5Kgqj2u12Nm49jh7ticR3MSb0RPbk5lewPuvrHYJgxiMMlff3JdRE9qzffNJvvxgo8ewT/c+zZh9XW/efG4FVosdgcCuOLjyht6MrUDxWm1QWGDlxUeXcCYtH4vZjsEgI8mCux4eTrdezRpcy0wtLVOj0eNJ6wbAbldwKA4kqXH2c9X7edP53hl0vndGrc6zokg6ouT6zmZV2L8nhdSUnGIpi8piMOoYOa6dS0emz97d4BbztloU5v+4m5Hj29Ek1AeH4r7QlGVB02b+RLcM4v2vZnLk4GkKC2106BxeXHldH3h5G3jh3UvYtTWBmEOnaRLuy+ARrasVYqovNIev0Sho3zmM/XtSOD/A27xFYK10yCqP/buT+fnbnaQkZhMQ5MWlV3ZneBnVtPXN8dgzHkMSOp1E0snsKjt8T5xKyvF4vCDfitlsp2kzf4SHSIosS4yd7FzFS7JEp25Na8ym6qLTSfQb3IJ+g2v+7qsu0dIyNRoF19zSDy+TvrjAR5IERqOOG++o+/j9oX2neP+V1SSccOZkp6fl88PnW1m2sOyuRvVJdMtgj/Fqxe6osoxEaTQJ9xzPNpr0GI06fvh8q8ec+kEjWxEe0TiyXRormsPXaBQ0iwrg5Q+mMm5yB9p1DGXk+Ha88O4U2nYMrXNbfvlhl8eQxR9z9+JQqlYcVNuMndwBnd71467XS7TrFOpR5rc6zLy6p1vHMoNRZursrjgcKpvXn/B4t7Fn+/mq6ho1jRbS0Wg0hIT6cNXNHvei6pSURM9FRVaLnYJ8G77+NRfbVVWVjDMFyLIgsBqiXiGhPjzx8gS+/2xLca77kFGtufqWmn89+wyM5ua7B/HLdzvJzCjAx9fItNldGT+1E1aL3WP8HsBcTu5+dUlJymbnlkRknaDfoBaNJrOmJtEcvkajJi/Xwk9fb2fbhnhUVaVn/yiuvbVfufrz8XEZzP1mB8eOpuHr61TpHHdJxwplXIQ19SM+LsPtuF4v4+VTc3oqJ46l88nb60k/kw+qSmR0IHc+NLzKYY8WrYN5+vVJOBQHQhK1ml0yaHgrBg1vhd2mIOuk4rmMJj1NI/1JTnD90hSCcpuOVIc/f97Dwt8P4FBUhAS//W831/6nn8tm88WAFtLRaLQ4FAcvPbaEzetOYLUq2GwOdmw6yfMPLyqzMjIlKZuXn1jKoX2nsFoUMtIL+PV/u5j33c4KzTvz6p6u+j6A3iDRvEUgzzzwD++8uJJD+05V67nl5Vp47enlnErOwVb03OLjMnj58SXYPYjJVQZJluoslVCnl93muunOgRiNuuJUWp1OwstbzxU31k5l8ckTmfzz+wFsVmcPWrvNgc2q8L/Pt5GVWVj+ABcQmsPXaLTs25VCZnoBSokNQIdDpSDPyo5NJ0u97u9f97lVh1otCisWHaEg31ruvD36RvKfEsU2Xt56JCGIi00nMT6LPTuSeOellaxdEVvFZwYb18S5PC9w6sFYLHZ2N/JYd/tOYbzw7hRGjm9Hxy7hTJjWiVfmTKNps5rdPD7L1vUnsNvdvySFBLu2JtTKnA0VLaSj0WhJSszyKJ1sNttJjM8EWnm8Li423WO5vE4ncfpULi3blC+93H9wC/oPblHcgnH1shiUErFpq0Xhp6+3M3hEqyqljZ45ne/xLsVud5CRno+iOMjPs+Lja0CWJVRVZeOa4yxdcIj8PAs9+zZn2uXdCAisemvF2qRpM39uuGNAla61Wuz88v1O1hZ1QuvQJYzrbutPZFRgqdc00PrSOkdz+BqNEofiwNfXiE4voyjntQQUsGVDPL5+RkZPao/R5BpXb9Y8gJSkHLecfrtNqfRGnk4nsX93iouzP4uqqpxKzqG5B02cxPhM4mLTCWniQ6duTd0qhdt1DGX1shi3RtuSEKSm5HLXdb9gsynodTJTZnYlP8/Cv4uPFrdHXLU0hm0b43nlg2k1uoncEPjgtdUc3p9aLLFxaH8qLz66hNc+nOZxY7v/kBYsXXDI7QtUdUCv/lF1YnNDQXP4Go2ODavj+Omr7VjMNueHXuDivM9K9M6fu4f1q+N49s3JLmmCU2d1Y9+u5GLhLgCDQabf4BZV6lwUEORFakqu23HF7sD3vPEUxcFHb65j384khBAICfz8TTzx8niCm5z7sunVP4qwpr6cSsopdmwGg0xIqLdTvbHIdrvNwV+/7EWxO1zuWhTFQUG+lX+XHGH65d0r/ZwaKskJ2Rw5cNpVTVN1atavWHyUWdf0dLsmulUwk2d04Z/fDxSnzaoqtGoXQuzhNHr1b47sQRb6QuTieJYaFwwH96bw7Sebycu1FH/oS9t+tFkV0lLz2LLuhMvxVm1DuO/xkYRH+CFJznaCIye04+a7qlbENfnSzm6buDqdRIeu4W7ZQssXHmbfriSsVgWLpahpSFo+n7y9zu36J1+dyMRLuxAa5kt4hB+XXtmdwkK7yxfV2efpKURlszk4sCelSs+poZKclO3ROdttDuLj0ku97rIre/DCO1MYNMIZ5hMCjh48zefvb+CVJ5Z6DA1eiGgrfI1Gxd+/7XdzeKrq3IAzGHRuIRCL2c6e7YkMG+PaKrBrz2a88cmlWCx29DoJqRorvF79o5hxVQ9+n7sHWZaw2x206xjKnQ8Oczt35dKjbvY7HCpxMenk5phd7jC8vPTMuqany6r1t//trrBdQhKEhjf+5uMlaRYZgOKhuE2nl2jRuvS9F1VVObI/lQ2r4lzi+RaznZMnMlm7IpYxkzrUhskNCs3hazQq0lLzPB7XyZLHnTlJEgSGlF6wZKwhLf1Jl3Zh9MT2JCVkExDoVepeQGldk4QQpT5WkohIf5IS3Au/zu4BlFzp63USE6Z2qoj59Y5DcbBnRxKH9qcSGOTFkFGtCQj0wmp1iuOd7RfrH2jCaNK5xeP1epkxk0rXpF/29yF+/WGXx81bq0Vh05rjmsPX0GhotOsYSnpavlsIQ0gCbx8jVmuBy4dap5MYXUfNKYwmPa3bld6WMC/HUmr+e1CIF0FlfDGd5cqb+jDntTXuGTxCRXU4QxV6g4zRqOPmuwaV24ikIWC1Krz+zDISTmRhMdvR62X++HkPLVoFExeTjopK8+hAbrl7EF9/tJmCfNeKXCEJ7nl0BEGlVCIrioM/5+1166JVkvNDchcqWgxfo1Ex/YruGIwyJf2mwSgz46oePPbSOMIj/DEaZUxeery89fzn/iE0i6pZrZiq8vaL/5KZ7t4o3GCUuf3+oRUqhureO5L/PjWKNu2buGT2OBTnDY6skxg5vj0ffDOL3gMaRwbKyiVHORmXWRyOs9kUrBaFmMNpKIoDh6Jy8ngmLz+xlJTEbLeQjiwJDpZR6Jafaynz7slo1DWajlXVRVvhVwJrdh4Ou4IppGE4kIuRps38efbNyfz+026OHDxNYJAXU2d1K+5N+tpH00hKyMZittGiVXC9SCd7IuFEJoknszxurvYZEF0pEbjO3SN44Klg7rv5N7fx7DYHu7YkcE0FNHKOxZzhs3fWc/pULrIs0X9oC265e3CFukDVJBtXHyuzMvosdpuCw8Mi3W53uEk1lMTb14gsS55X+AKGjGpF30HRlTG50aI5/AqQn5TG2utf4/SG/QAEdIhm2HePEtKz4eqfX8g0ax7A3Y+M8PiYEILm0YF1a1AFcAqgSYC7Y8vOqnx5v+JwlJqeZK+AYmfiySxefGRxcfjLbnewcfVx4o9l8MqcaZW2pzpUNCXSk7M/S3CT0sNhOp3E5Mu6sPB31w1/WSdxwx0DGDH24vkcayGdcnAoCouG/5fUtXtxWO04rHYy98WxeOR/MZ8pfVWhoVGS6NZBpab+RbUMrPR4AYFehDd1F1HT6SQGDC2/Scc3H23yuIGZlJDNkQPV0wGqLKMmtK/Q5rlUhrcKDSs7G2na5d249MoexZ2zmoT5cOeDQy8qZw+awy+XlBU7MZ/JQj1v1eSwKcR8u6SerNJoCKiqSmGhzWOY5nyCgr3p1a+5x8d2bk6kKr2lb//vULy89cVFZUaTjtBwX6bNLr/QKiE+q9THFvy2v9K2VIeho1rTo28kBoOMTi9h8tIhy8IltCRJAr1RRq93d1lGk0xkOXd1QgimXNaFj364nK9+u4a3P59B30GNu3tVVdBCOuWQe+KUm7MHUAot5MQm14NFGg2BDavjmPfdTvJyzBiMOiZd2pmps7qV2UzdYNS5VQUD5OSYiYs5Q5v2lWvmojfI9B/aktjDp/H1MzJsdBsGDa+Ydo/J5F6zcJbSUl9rC0mWuOvh4cTHZXD04Gn8A01069WMJX8dZPXyWGxWOz37NueSWV156bGl2OzW4tdQkgWBQd506VExaWUhBDpdw2o6XpdoDr8cQnq3Q3gIlup8vQgb1LkeLNKob3ZuSeDbjzcXbzQWFthYOH8/qkPl0it7lHpddmahm7MH5+o1N8dSKRt2b0vkozfXYi+SVDAYZbIyCukzMLpCDn/8tE78+v0uj49F1JJqZXm0aB1Mi9bONNLEk1n4B3px7a396NmvefFdzFOvTeCLDzYSH5eBALr0jOCWuweX+UWrcQ7N4ZdDaL+OhA3uTOqG/SiFTulcyaDDFBpIq8tH1q9xGvXC73N3e2xxuPjPg0yd3a3UTcie/Zpz9OBpt2ttNoU27UvP3z8fRXHwxQcbXcaxWhTSz+SzZMFBZlzVs9wxLpnRlcV/HCAv11UOWqcXjK/HYi2HQ+XLORudDW0AWRbIssRjL44julUwkVGBPPfmZMyFtiJZjIbhwpITs9m09jg2q0LfgdG06dCkznoOVAYthl8Bxv79Mt0evQqfqFBM4UG0v3UKU7d+hM7rwlIh1KgYZ07nezxutzsozC+9Td/wMW0ICfVBX0LIzWjUMW1Wt0qJtiUnZHtsgmK3Odi2sfQ+AOfzwjuXENHcH71ewmiS0RtkZl3Tq8Lhkdpg87rjbN940tnQxqpgLrSTn2flvZdXuexzmLz0DcbZr1h0hGce+IeF8/ez+M+DvP7scr79ZEuV9mVqm4bxijVwZKOBXs9cT69nrq9vUzQaAJFRgcQeSXM7bvLS412UBeIJo0nPs29NZuWSo2zfdBJfXwNjp3SkR5/ISs1v8tKjlLJRnJdrweFQKxTiCAn14dU500g4kUlujoVWbUPw9ind/oqSlVnIgl/2smdHEj4+BiZM68zgka0qtOJdvSwGi8V9byEvz8rJ45nFIZ+GQnZWIT9/s90lx/+sVMOQka1p3zmsHq1zR3P4GhqVZPb1vXj7+X9dQioGo8ysa3uW62i9vPRMuawLUy7rUuX5Q8N9adY8gPjjGW57Avl5Fpb8dZDJFRxfCFEl+QVVVbHZHOj1ru0S83IsPPPfheTlWlAUlTPk8+2nW0iIz+TKG/uUO67d7jnZXgg8iqbVN3t2JCFJEuBqm8VqZ8v6Ew3O4VcrpCOECBZCLBdCxBT9797pwXmeIoTYXfRvQXXm1NCobzp2CefBZ8bQun0TjEYdTZv5c+s9g+u0PP+eR4Z73ABW7CpL/z5Ua/OqqsqiPw9w13W/cNsVc7n/lvlsXBNX/Pi/i49QkG87r/uXnRX/HCYn21zu+INHtPKoayPrpAa3uoeiojEP3/ECPKaQ1jfVXeE/BvyrquprQojHin5/1MN5haqq9qzmXBoaDYaOXcN59o1J9TZ/QJAXQnhu3VeQV35f3qryzx8H+Gve3uKK1ayMQj57dwM7Np/k/x4YxoG9KR4LzHR6mfi4DLr1albm+CPGtWPL+nji4zKwmO3o9BKSJLjzwWENsklJz77N+dax2e24Xi8zaETrerCobKrr8KcDI4t+/g5YjWeHr6FxUZOUkMXSBYc4lZxDhy7hjJvcAf9q9Js1GHU0jfQnJTHH7bF2nSqXz19RHA6VhR76EQBs35TAnNfXlKrDoyiOCqmB6vUyj784jr07k9m/O5mAIC+Gjm5TqhJmfePja+D/HhzKJ2+vR0jCWYSnwvQruzfIOxJRnZ1kIUSWqqqBRT8LIPPs7+edZwd2A3bgNVVV/yxlvNuA2wCio6P7xMfHV9k2DY2Gwv7dybz/6mrsNmfOvE4vYTLpeeGdKZXuoVuSQ/tO8c5LK7HZHKhFG7V6g8zTr00kqqXH6Gq1KCywctd1v3js3wvOOLtO5y5SJsuCFq2DefbNyTVuU0MhL9fCzq0J2G0OevSJrNbftboIIXaoqupRPa9chy+EWAE09fDQk8B3JR28ECJTVVW3d5oQIlJV1SQhRGtgJTBGVdVjZc3bt29fdfv27WXapqHR0FFVlQdv+4P0NNdUTiFg8MjW3HbfkGqNn3Aik39+P0BSQhat24Uw+bKuhEe4a+zUBA6Hyj03/EpebuWKxDp2DePuR0ZUqV+wRuUpy+GXG9JRVXVsGQOnCiEiVFVNEUJEAKdLGSOp6P84IcRqoBdQpsPX0LgQyM4sJMeDGqaqwqa1x7nmlr74+Fa9niOqZRB3PDC0zHPyciyknsolNMynWmEkSRLMurYn33+2pUzlypKYTDrGTemkOfsGQnVj+AuAG4DXiv7/6/wTijJ3ClRVtQghmgBDgDeqOa+GRqPA6KX3uLEK4FBUvpyzifseH1krczsUBz98uY21K2LR62VsNoUBQ1py892Dqqx5P2pCe+x2hR+/3O7yvErbQEYIdA0wW+Vipbp/ideAcUKIGGBs0e8IIfoKIb4sOqcTsF0IsQdYhTOGf7Ca82poNAq8vPR07116YdXeHUkeC41qgn/+OMD6lcew2xwUFtiw2xxs3RjPb//zrKFTUcZN6cTzb0+hVduQIjEyia49Ior1bkoihLNhi0bDoFqbtrWJFsO/uHEoDrasj2fD6jh0eokRY9vSs1/zBqlPUh75eVbuvv4XjzLKsk7ig29m4etX8zId99zwq8fcd6NJx2dzr6yR19JmU5AlgSRL/D53D4t+P4CQBFJRT/kHnhpNx67h1Z5Ho+JUK4avoVHXqKrKe6+s5vD+1OLV78E9pxgyqjU33DGgnq2rPD6+BgYOa8mmtcfdwh5h4b614uwBCvI95+NbLHYcDhVZrr7D15dQ5pxxVQ9GjG3Lvt3JmEw6evZtjslLX+05NGoOLbim0eA4sCeFwwdSXUIdFouddSuPldm7tCEz+7pe+Pobi4XTZJ2E0aTjlnsG1dqcrUtR4GweHVhrRUwhoT6MHNeOgcNaac6+AaKt8DUaHHt3JntuzqHCgb0pNItqfE3kg5v48NqH01m9LIaYQ6eJiPRnzOSOhIaX3ZqvOlx9c19efXIZNpuCw6EiJIFeL3H9bf1rbc76ID/PyprlMRzan0rTCD/GTulAeET9aPo3dDSHr9Hg8PUzoNNL2D0U8PiUoUbZ0PH1M3LJzK51Nl+rtiE8/85kFs7fz4ljGTSPDuSSmV1rpCgrOTGbM6fziGoZVK9VsFmZhTz7wD/k51uxWRUOyILVy2N44KnRdOrmqXzo4kZz+BoNjiEj2/D3rx76qgro3T+q7g1qxEREBvCfe6tX3FWSgnwr772yiuMx6cg6CbtNYcioNtxwx4B66Tr1x9w95OaYi6t/FUVFURS+nLORtz67rFFu8tcmWgxfo8EREurD/z00DJOXDi8vPSYvHb5+Rh56dowWF65nvvpoE8eOnsFqVSgssGGzOdi4Jo4V/xyuF3t2b0/0KPWQnWUmM8O94O1iR1vhazRIevePYs53lxNz6DSyTqJdx9AGqZZ4MWEx29i9NdFNs95qUVj2z+F6aY1oMnl2YapDxdhAOmI1JLRPkEaDxWCQ6dIjgo5dwi8IZ5+els+hfafI9iC10BiweFDJPEtZrR1rk7FTOnrUzwf47cddFdLgv5jQvgI1NGoYVVUpLLChN8jo9TJWi52P317H/l0p6PQSNpvCsFFtuP72/ki1/EVmtSo1dpfk528kKMSbtNQ8l+NCQNde9VNNO2ZSB+LjMti0Jg5FUYvrHOx2B2uWxbJ7WyKvzJmGlxYKBDSHr6FRoxzen8rXH2/iTGoeQhL0H9wCSRLs3+1sDHK2OciGNXGERvhVq9VheezcksBn761HIFBR0elk7n9yJO06Vq3tnhCCm+8ayLsvr8ZelOrplHrWMfvaXjVsfUVtOvfD+aoBiuIgL9fC+pXHGDelY90b1wDRpBU0NGqI5MRsnntwkUvBmE4vodgdHoXFgkK8ee+rmbViS3paPo/e9Rc2q2sYxuSl54NvZmI0VX3Fm5yY7WzmkpRD+y5h1W7mUh22bYzni/c3lqlH1HdQNPc8OqIOrapfNGkFDY06YMlfB93a+51fS1CS0qQPaoINq4+hetDuUVWVnVsSGTSiVZXHbtY8gJvuHFgd82qMNStiy3T2Op1E02a10x+gMdL4d8I0NBoIyYnZHgXSPKaCC+jQpWqhlYqQl2t1y6YBpyRzfi1+0dQ1iofnWBJZluq0uXxDR3P4Gho1RLuOYR515iVZQm+QiguTZJ2El0nPVTd6vOuuEbr3bobRU8qigC49LpwK1KGj2pSafhka5stDz46hSVjtyVc0NrSQjoZGDTF+akdWLzuKopyL2RuMMgOHtWLi9E4s/uMgSQlZtOkQysRpnWrVEXXu7kxnLak4ajTqGDKqNRGRjU+LqDQGDm/JlvUnnGJ7Zjs6vYQQcNOdgxg8opVWaXse2qathkYNkpqSw7zvdnFwbwre3nrGTe3EhEs61nr6pScupJ4CZaGqKgf3nmL/7mT8/E0MGtGqXvV96ptqNTGvLzSHr6GhoVF5ynL4WgxfQ+MiQFVVkhOySTiR6XFjWePiQIvha2hc4Jw8kckHr64mO6sQIQReXnrueng47TvXXpaQRsNEW+FraFzAWCx2XntqGWmpeVgtChaznazMQt564V9NZ+YiRHP4GhoXMDu3JHjMVXc4VDatOV4PFmnUJ5rD19C4gMnOKvRYgGWzKmRlFtSDRRr1iebwNTQuYDp0DkeS3dMwjSYdHbtcOAVYGhVDc/gaGhcwrdqG0K1XM4wlNOMNRpnoVkF0qydJY436Q8vS0dC4wLn74eGs/TeW1ctjcSgOhoxqw+iJ7eulGEyjftEcvobGBY4kS4wc356R4zURsYsd7SteQ0ND4yJBc/gaGhoaFwlaSEdDQ6POSE/LZ93KY2RlFNClRwS9B0RdEA3qGwuaw9fQ0KgT9u9O5v1XV+NwqNhtDjauOU6zqACeeGk8hlI07TVqFu2rVUNDo9ZxKA4+eWc9VotS3PbRYraTFJ/FyiVH69m6iwfN4WtoaNQ6CfFZ2M9rqA5gtSps1CQe6gzN4WtoaNQ6Or1Eab03DAbZ43GNmkdz+BoaGrVOs+YBBAR5w3kqD0ajTmsyXodoDl9DQ6PWEUJw/5Mj8fMzYvLSYTDKGAwyfQdHM2hEq/o276KhWlvjQojZwHNAJ6C/qqoeexIKISYC7wMy8KWqqq9VZ14NDY3GR2RUIO99NZO9O5PJziqkQ+dwmkVdOA3VGwPVzYXaD8wAPivtBCGEDHwEjAMSgW1CiAWqqh6s5twaGhqNDJ1epveAqPo246KlWg5fVdVD4LxdK4P+QKyqqnFF5/4MTAc0h6+hoaFRh9RFDD8SSCjxe2LRMTeEELcJIbYLIbanpaXVgWkaGhoaFw/lrvCFECsAT50SnlRV9a+aNEZV1c+BzwH69u3rOYdLQ0NDQ6NKlOvwVVUdW805koCSQbvmRcc0NDQ0NOqQugjpbAPaCSFaCSEMwJXAgjqYV0NDQ0OjBKK06rcKXSzEZcAcIBTIAnarqjpBCNEMZ/rl5KLzJgPv4UzL/FpV1ZcrMHYaEF/0axPgTJUNrVsai62anTWLZmfN0ljshIZnawtVVUM9PVAth19XCCG2q6rat77tqAiNxVbNzppFs7NmaSx2QuOyVau01dDQ0LhI0By+hoaGxkVCY3H4n9e3AZWgsdiq2VmzaHbWLI3FTmhEtjaKGL6GhoaGRvVpLCt8DQ0NDY1qojl8DQ0NjYuEBunwhRCzhRAHhBAOIUSp6U5CiBNCiH1CiN1CCI/SzLVNJWydKIQ4IoSIFUI8Vpc2Fs0fLIRYLoSIKfo/qJTzlKLXc7cQos4K5Mp7fYQQRiHEvKLHtwghWtaVbefZUZ6dNwoh0kq8hrfWg41fCyFOCyH2l/K4EEJ8UPQc9gohete1jUV2lGfnSCFEdonX8pm6trHIjighxCohxMGiz/p9Hs5pEK9puaiq2uD+4dTX7wCsBvqWcd4JoElDtxVnwdkxoDVgAPYAnevYzjeAx4p+fgx4vZTz8urhNSz39QHuBD4t+vlKYF4DtfNG4MO6tu08G4YDvYH9pTw+GViMs//UQGBLA7VzJLCwPl/LIjsigN5FP/sBRz383RvEa1revwa5wldV9ZCqqkfq246KUEFbiyWiVVW1AmclouuS6cB3RT9/B1xax/OXRUVen5L2/waMEeXoctcCDeHvWC6qqq4FMso4ZTrwvepkMxAohIioG+vOUQE7GwSqqqaoqrqz6Odc4BDuir8N4jUtjwbp8CuBCiwTQuwQQtxW38aUQYUlomuRcFVVU4p+PgWEl3KeqUiierMQ4tK6Ma1Cr0/xOaqq2oFsIKROrPNgQxGl/R1nFt3W/yaEaIjdPhrC+7GiDBJC7BFCLBZCdKlvY4pCib2ALec91Che0+p2vKoyNSS7PFRV1SQhRBiwXAhxuGjVUKPUpUR0dSjLzpK/qKqqCiFKy8dtUfSatgZWCiH2qap6rKZtvYD5G5irqqpFCHE7zruS0fVsU2NlJ873Y16RHtefQLv6MkYI4QvMB+5XVTWnvuyoDvXm8NXqyy6jqmpS0f+nhRB/4LzlrnGHXwO21olEdFl2CiFShRARqqqmFN1qni5ljLOvaZwQYjXO1UxtO/yKvD5nz0kUQuiAACC9lu06n3LtVFW1pE1f4tw7aWg0Csnykk5VVdVFQoiPhRBNVFWtc6EyIYQep7P/UVXV3z2c0ihe00Yb0hFC+Agh/M7+DIzH2WO3IdIQJKIXADcU/XwD4HZnIoQIEkIYi35uAgyhblpRVuT1KWn/LGClWrRbVoeUa+d5cdtpOOO9DY0FwPVFmSUDgewS4b4GgxCi6dl9GiFEf5z+qq6/5Cmy4SvgkKqq75RyWqN4Tet919jTP+AynDEwC5AKLC063gxYVPRza5xZEnuAAzjDKw3SVvXcLv5RnKvlOrcVZ7z7XyAGWAEEFx3vi1PKGmAwsK/oNd0H3FKH9rm9PsALwLSin03Ar0AssBVoXU9/7/LsfLXo/bgHWAV0rAcb5wIpgK3ovXkLcAdwR9HjAvio6Dnso4xMuHq28+4Sr+VmYHA92TkU537hXmB30b/JDfE1Le+fJq2goaGhcZHQaEM6GhoaGhqVQ3P4GhoaGhcJmsPX0NDQuEjQHL6GhobGRYLm8DU0NDQuEjSHr6GhoXGRoDl8DQ0NjYuE/wdWvWsfZiTFMAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot data\n",
    "y_pred = np.argmax(nn.z2, axis=1)\n",
    "\n",
    "plt.scatter(X[:, 0], X[:, 1], c=nn.y, cmap=plt.cm.Spectral)\n",
    "plt.title(\"ground truth\")\n",
    "plt.show()\n",
    "\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y_pred, cmap=plt.cm.Spectral)\n",
    "plt.title(\"predicted\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 9. 如何使用类的方法封装多层神经网络?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn import datasets, linear_model\n",
    "from sklearn.metrics import accuracy_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "# define sigmod\n",
    "def sigmod(X):\n",
    "    return 1.0/(1+np.exp(-X))\n",
    "\n",
    "\n",
    "# generate the NN model\n",
    "class NN_Model:\n",
    "    def __init__(self, nodes=None):\n",
    "        self.epsilon = 0.01                 # learning rate\n",
    "        self.n_epoch = 1000                 # iterative number\n",
    "        \n",
    "        if not nodes:\n",
    "            self.nodes = [2, 6, 2]          # default nodes size (from input -> output)\n",
    "        else:\n",
    "            self.nodes = nodes\n",
    "    \n",
    "    def init_weight(self):\n",
    "        W = []\n",
    "        B = []\n",
    "        \n",
    "        n_layer = len(self.nodes)\n",
    "        for i in range(n_layer-1):\n",
    "            w = np.random.randn(self.nodes[i], self.nodes[i+1]) / np.sqrt(self.nodes[i])\n",
    "            b = np.random.randn(1, self.nodes[i+1])\n",
    "            \n",
    "            W.append(w)\n",
    "            B.append(b)\n",
    "            \n",
    "        self.W = W\n",
    "        self.B = B\n",
    "    \n",
    "    def forward(self, X):\n",
    "        Z = []\n",
    "        x0 = X\n",
    "        for i in range(len(self.nodes)-1):\n",
    "            z = sigmod(np.dot(x0, self.W[i]) + self.B[i])\n",
    "            x0 = z\n",
    "            \n",
    "            Z.append(z)\n",
    "        \n",
    "        self.Z = Z\n",
    "        return Z[-1]\n",
    "        \n",
    "    # back-propagation\n",
    "    def backpropagation(self, X, y, n_epoch=None, epsilon=None):\n",
    "        if not n_epoch: n_epoch = self.n_epoch\n",
    "        if not epsilon: epsilon = self.epsilon\n",
    "        \n",
    "        self.X = X\n",
    "        self.Y = y\n",
    "        \n",
    "        for i in range(n_epoch):\n",
    "            # forward to calculate each node's output\n",
    "            self.forward(X)\n",
    "\n",
    "            self.evaluate()\n",
    "            \n",
    "            # calc weights update\n",
    "            W = self.W\n",
    "            B = self.B\n",
    "            Z = self.Z\n",
    "            \n",
    "            D = []\n",
    "            d0 = y\n",
    "            n_layer = len(self.nodes)\n",
    "            for j in range(n_layer-1, 0, -1):\n",
    "                jj = j - 1\n",
    "                z = self.Z[jj]\n",
    "                \n",
    "                if j == n_layer - 1:\n",
    "                    d = z*(1-z)*(d0 - z)\n",
    "                else:\n",
    "                    d = z*(1-z)*np.dot(d0, W[j].T)\n",
    "                    \n",
    "                d0 = d\n",
    "                D.insert(0, d)\n",
    "            \n",
    "            # update weights\n",
    "            for j in range(n_layer-1, 0, -1):\n",
    "                jj = j - 1\n",
    "                \n",
    "                if jj != 0:\n",
    "                    W[jj] += epsilon * np.dot(Z[jj-1].T, D[jj])\n",
    "                else:\n",
    "                    W[jj] += epsilon * np.dot(X.T, D[jj])\n",
    "                    \n",
    "                B[jj] += epsilon * np.sum(D[jj], axis=0)\n",
    "        \n",
    "    def evaluate(self):\n",
    "        z = self.Z[-1]\n",
    "        \n",
    "        # print loss, accuracy\n",
    "        L = np.sum((z - self.Y)**2)\n",
    "            \n",
    "        y_pred = np.argmax(z, axis=1)\n",
    "        y_true = np.argmax(self.Y, axis=1)\n",
    "        acc = accuracy_score(y_true, y_pred)\n",
    "        \n",
    "        print(\"L = %f, acc = %f\" % (L, acc))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# generate sample data\n",
    "np.random.seed(0)\n",
    "X, y = datasets.make_moons(200, noise=0.20)\n",
    "\n",
    "# generate nn output target\n",
    "t = np.zeros((X.shape[0], 2))\n",
    "t[np.where(y==0), 0] = 1\n",
    "t[np.where(y==1), 1] = 1\n",
    "\n",
    "# plot data\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y, cmap=plt.cm.Spectral)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 121.621107, acc = 0.500000\n",
      "L = 115.928422, acc = 0.500000\n",
      "L = 111.304997, acc = 0.500000\n",
      "L = 107.789222, acc = 0.500000\n",
      "L = 105.265297, acc = 0.500000\n",
      "L = 103.533617, acc = 0.500000\n",
      "L = 102.380546, acc = 0.500000\n",
      "L = 101.622557, acc = 0.500000\n",
      "L = 101.121698, acc = 0.500000\n",
      "L = 100.782803, acc = 0.510000\n",
      "L = 100.543751, acc = 0.530000\n",
      "L = 100.365372, acc = 0.540000\n",
      "L = 100.223492, acc = 0.520000\n",
      "L = 100.103371, acc = 0.475000\n",
      "L = 99.996073, acc = 0.460000\n",
      "L = 99.896185, acc = 0.465000\n",
      "L = 99.800411, acc = 0.465000\n",
      "L = 99.706725, acc = 0.495000\n",
      "L = 99.613854, acc = 0.515000\n",
      "L = 99.520981, acc = 0.560000\n",
      "L = 99.427551, acc = 0.585000\n",
      "L = 99.333171, acc = 0.630000\n",
      "L = 99.237541, acc = 0.660000\n",
      "L = 99.140415, acc = 0.690000\n",
      "L = 99.041582, acc = 0.705000\n",
      "L = 98.940844, acc = 0.710000\n",
      "L = 98.838015, acc = 0.720000\n",
      "L = 98.732913, acc = 0.740000\n",
      "L = 98.625357, acc = 0.745000\n",
      "L = 98.515164, acc = 0.755000\n",
      "L = 98.402148, acc = 0.785000\n",
      "L = 98.286120, acc = 0.790000\n",
      "L = 98.166887, acc = 0.800000\n",
      "L = 98.044250, acc = 0.800000\n",
      "L = 97.918005, acc = 0.805000\n",
      "L = 97.787942, acc = 0.815000\n",
      "L = 97.653845, acc = 0.830000\n",
      "L = 97.515489, acc = 0.830000\n",
      "L = 97.372644, acc = 0.830000\n",
      "L = 97.225071, acc = 0.830000\n",
      "L = 97.072523, acc = 0.830000\n",
      "L = 96.914745, acc = 0.835000\n",
      "L = 96.751472, acc = 0.835000\n",
      "L = 96.582430, acc = 0.835000\n",
      "L = 96.407335, acc = 0.835000\n",
      "L = 96.225894, acc = 0.835000\n",
      "L = 96.037800, acc = 0.835000\n",
      "L = 95.842740, acc = 0.835000\n",
      "L = 95.640384, acc = 0.835000\n",
      "L = 95.430396, acc = 0.835000\n",
      "L = 95.212423, acc = 0.835000\n",
      "L = 94.986104, acc = 0.830000\n",
      "L = 94.751064, acc = 0.830000\n",
      "L = 94.506915, acc = 0.830000\n",
      "L = 94.253259, acc = 0.830000\n",
      "L = 93.989683, acc = 0.830000\n",
      "L = 93.715765, acc = 0.830000\n",
      "L = 93.431069, acc = 0.830000\n",
      "L = 93.135151, acc = 0.830000\n",
      "L = 92.827554, acc = 0.830000\n",
      "L = 92.507814, acc = 0.830000\n",
      "L = 92.175457, acc = 0.830000\n",
      "L = 91.830004, acc = 0.835000\n",
      "L = 91.470973, acc = 0.835000\n",
      "L = 91.097875, acc = 0.835000\n",
      "L = 90.710225, acc = 0.840000\n",
      "L = 90.307539, acc = 0.845000\n",
      "L = 89.889339, acc = 0.845000\n",
      "L = 89.455160, acc = 0.845000\n",
      "L = 89.004546, acc = 0.840000\n",
      "L = 88.537066, acc = 0.840000\n",
      "L = 88.052308, acc = 0.840000\n",
      "L = 87.549895, acc = 0.840000\n",
      "L = 87.029483, acc = 0.845000\n",
      "L = 86.490773, acc = 0.845000\n",
      "L = 85.933518, acc = 0.845000\n",
      "L = 85.357526, acc = 0.845000\n",
      "L = 84.762674, acc = 0.845000\n",
      "L = 84.148911, acc = 0.845000\n",
      "L = 83.516272, acc = 0.845000\n",
      "L = 82.864878, acc = 0.845000\n",
      "L = 82.194952, acc = 0.845000\n",
      "L = 81.506820, acc = 0.840000\n",
      "L = 80.800921, acc = 0.840000\n",
      "L = 80.077810, acc = 0.840000\n",
      "L = 79.338167, acc = 0.840000\n",
      "L = 78.582791, acc = 0.840000\n",
      "L = 77.812612, acc = 0.840000\n",
      "L = 77.028680, acc = 0.840000\n",
      "L = 76.232171, acc = 0.840000\n",
      "L = 75.424374, acc = 0.840000\n",
      "L = 74.606691, acc = 0.840000\n",
      "L = 73.780620, acc = 0.840000\n",
      "L = 72.947751, acc = 0.840000\n",
      "L = 72.109745, acc = 0.840000\n",
      "L = 71.268324, acc = 0.840000\n",
      "L = 70.425252, acc = 0.840000\n",
      "L = 69.582316, acc = 0.840000\n",
      "L = 68.741307, acc = 0.840000\n",
      "L = 67.904004, acc = 0.840000\n",
      "L = 67.072151, acc = 0.840000\n",
      "L = 66.247442, acc = 0.840000\n",
      "L = 65.431502, acc = 0.840000\n",
      "L = 64.625872, acc = 0.840000\n",
      "L = 63.831996, acc = 0.840000\n",
      "L = 63.051206, acc = 0.840000\n",
      "L = 62.284717, acc = 0.840000\n",
      "L = 61.533617, acc = 0.840000\n",
      "L = 60.798864, acc = 0.840000\n",
      "L = 60.081280, acc = 0.840000\n",
      "L = 59.381556, acc = 0.840000\n",
      "L = 58.700250, acc = 0.840000\n",
      "L = 58.037794, acc = 0.840000\n",
      "L = 57.394496, acc = 0.840000\n",
      "L = 56.770551, acc = 0.840000\n",
      "L = 56.166043, acc = 0.840000\n",
      "L = 55.580959, acc = 0.840000\n",
      "L = 55.015197, acc = 0.840000\n",
      "L = 54.468573, acc = 0.840000\n",
      "L = 53.940833, acc = 0.840000\n",
      "L = 53.431659, acc = 0.840000\n",
      "L = 52.940684, acc = 0.840000\n",
      "L = 52.467494, acc = 0.840000\n",
      "L = 52.011639, acc = 0.840000\n",
      "L = 51.572642, acc = 0.840000\n",
      "L = 51.150004, acc = 0.840000\n",
      "L = 50.743209, acc = 0.840000\n",
      "L = 50.351731, acc = 0.840000\n",
      "L = 49.975042, acc = 0.840000\n",
      "L = 49.612610, acc = 0.835000\n",
      "L = 49.263906, acc = 0.835000\n",
      "L = 48.928410, acc = 0.840000\n",
      "L = 48.605606, acc = 0.840000\n",
      "L = 48.294993, acc = 0.840000\n",
      "L = 47.996079, acc = 0.840000\n",
      "L = 47.708390, acc = 0.840000\n",
      "L = 47.431462, acc = 0.840000\n",
      "L = 47.164849, acc = 0.840000\n",
      "L = 46.908123, acc = 0.840000\n",
      "L = 46.660868, acc = 0.840000\n",
      "L = 46.422687, acc = 0.840000\n",
      "L = 46.193200, acc = 0.840000\n",
      "L = 45.972040, acc = 0.840000\n",
      "L = 45.758860, acc = 0.840000\n",
      "L = 45.553325, acc = 0.840000\n",
      "L = 45.355116, acc = 0.840000\n",
      "L = 45.163929, acc = 0.835000\n",
      "L = 44.979474, acc = 0.835000\n",
      "L = 44.801473, acc = 0.835000\n",
      "L = 44.629662, acc = 0.835000\n",
      "L = 44.463789, acc = 0.835000\n",
      "L = 44.303614, acc = 0.835000\n",
      "L = 44.148907, acc = 0.835000\n",
      "L = 43.999451, acc = 0.835000\n",
      "L = 43.855036, acc = 0.835000\n",
      "L = 43.715465, acc = 0.835000\n",
      "L = 43.580546, acc = 0.835000\n",
      "L = 43.450099, acc = 0.835000\n",
      "L = 43.323950, acc = 0.835000\n",
      "L = 43.201935, acc = 0.835000\n",
      "L = 43.083894, acc = 0.835000\n",
      "L = 42.969678, acc = 0.835000\n",
      "L = 42.859141, acc = 0.835000\n",
      "L = 42.752145, acc = 0.835000\n",
      "L = 42.648557, acc = 0.835000\n",
      "L = 42.548251, acc = 0.835000\n",
      "L = 42.451106, acc = 0.835000\n",
      "L = 42.357004, acc = 0.835000\n",
      "L = 42.265834, acc = 0.835000\n",
      "L = 42.177489, acc = 0.835000\n",
      "L = 42.091866, acc = 0.845000\n",
      "L = 42.008866, acc = 0.845000\n",
      "L = 41.928395, acc = 0.845000\n",
      "L = 41.850363, acc = 0.845000\n",
      "L = 41.774680, acc = 0.845000\n",
      "L = 41.701264, acc = 0.845000\n",
      "L = 41.630034, acc = 0.845000\n",
      "L = 41.560912, acc = 0.845000\n",
      "L = 41.493823, acc = 0.845000\n",
      "L = 41.428697, acc = 0.845000\n",
      "L = 41.365463, acc = 0.845000\n",
      "L = 41.304056, acc = 0.850000\n",
      "L = 41.244412, acc = 0.850000\n",
      "L = 41.186469, acc = 0.850000\n",
      "L = 41.130168, acc = 0.850000\n",
      "L = 41.075452, acc = 0.850000\n",
      "L = 41.022266, acc = 0.850000\n",
      "L = 40.970558, acc = 0.850000\n",
      "L = 40.920276, acc = 0.850000\n",
      "L = 40.871372, acc = 0.850000\n",
      "L = 40.823798, acc = 0.850000\n",
      "L = 40.777509, acc = 0.850000\n",
      "L = 40.732461, acc = 0.855000\n",
      "L = 40.688613, acc = 0.855000\n",
      "L = 40.645922, acc = 0.855000\n",
      "L = 40.604351, acc = 0.855000\n",
      "L = 40.563861, acc = 0.855000\n",
      "L = 40.524415, acc = 0.855000\n",
      "L = 40.485980, acc = 0.855000\n",
      "L = 40.448521, acc = 0.855000\n",
      "L = 40.412004, acc = 0.855000\n",
      "L = 40.376400, acc = 0.855000\n",
      "L = 40.341678, acc = 0.855000\n",
      "L = 40.307807, acc = 0.855000\n",
      "L = 40.274761, acc = 0.855000\n",
      "L = 40.242511, acc = 0.855000\n",
      "L = 40.211032, acc = 0.855000\n",
      "L = 40.180297, acc = 0.855000\n",
      "L = 40.150284, acc = 0.855000\n",
      "L = 40.120967, acc = 0.855000\n",
      "L = 40.092325, acc = 0.855000\n",
      "L = 40.064334, acc = 0.855000\n",
      "L = 40.036975, acc = 0.855000\n",
      "L = 40.010226, acc = 0.855000\n",
      "L = 39.984068, acc = 0.855000\n",
      "L = 39.958481, acc = 0.855000\n",
      "L = 39.933446, acc = 0.855000\n",
      "L = 39.908947, acc = 0.855000\n",
      "L = 39.884966, acc = 0.855000\n",
      "L = 39.861486, acc = 0.855000\n",
      "L = 39.838490, acc = 0.855000\n",
      "L = 39.815964, acc = 0.855000\n",
      "L = 39.793892, acc = 0.855000\n",
      "L = 39.772260, acc = 0.855000\n",
      "L = 39.751053, acc = 0.855000\n",
      "L = 39.730259, acc = 0.855000\n",
      "L = 39.709863, acc = 0.855000\n",
      "L = 39.689852, acc = 0.855000\n",
      "L = 39.670216, acc = 0.855000\n",
      "L = 39.650941, acc = 0.855000\n",
      "L = 39.632017, acc = 0.855000\n",
      "L = 39.613431, acc = 0.855000\n",
      "L = 39.595173, acc = 0.855000\n",
      "L = 39.577233, acc = 0.855000\n",
      "L = 39.559600, acc = 0.855000\n",
      "L = 39.542265, acc = 0.855000\n",
      "L = 39.525218, acc = 0.855000\n",
      "L = 39.508449, acc = 0.855000\n",
      "L = 39.491950, acc = 0.855000\n",
      "L = 39.475713, acc = 0.855000\n",
      "L = 39.459727, acc = 0.855000\n",
      "L = 39.443987, acc = 0.855000\n",
      "L = 39.428483, acc = 0.855000\n",
      "L = 39.413208, acc = 0.855000\n",
      "L = 39.398154, acc = 0.855000\n",
      "L = 39.383314, acc = 0.855000\n",
      "L = 39.368682, acc = 0.855000\n",
      "L = 39.354250, acc = 0.855000\n",
      "L = 39.340012, acc = 0.855000\n",
      "L = 39.325961, acc = 0.860000\n",
      "L = 39.312091, acc = 0.860000\n",
      "L = 39.298397, acc = 0.860000\n",
      "L = 39.284872, acc = 0.860000\n",
      "L = 39.271510, acc = 0.860000\n",
      "L = 39.258306, acc = 0.860000\n",
      "L = 39.245255, acc = 0.860000\n",
      "L = 39.232351, acc = 0.860000\n",
      "L = 39.219590, acc = 0.860000\n",
      "L = 39.206966, acc = 0.860000\n",
      "L = 39.194474, acc = 0.860000\n",
      "L = 39.182111, acc = 0.860000\n",
      "L = 39.169870, acc = 0.860000\n",
      "L = 39.157749, acc = 0.860000\n",
      "L = 39.145742, acc = 0.860000\n",
      "L = 39.133846, acc = 0.850000\n",
      "L = 39.122056, acc = 0.850000\n",
      "L = 39.110369, acc = 0.850000\n",
      "L = 39.098780, acc = 0.850000\n",
      "L = 39.087286, acc = 0.850000\n",
      "L = 39.075884, acc = 0.850000\n",
      "L = 39.064569, acc = 0.850000\n",
      "L = 39.053338, acc = 0.850000\n",
      "L = 39.042188, acc = 0.850000\n",
      "L = 39.031116, acc = 0.850000\n",
      "L = 39.020118, acc = 0.850000\n",
      "L = 39.009191, acc = 0.850000\n",
      "L = 38.998332, acc = 0.850000\n",
      "L = 38.987539, acc = 0.850000\n",
      "L = 38.976808, acc = 0.850000\n",
      "L = 38.966136, acc = 0.850000\n",
      "L = 38.955522, acc = 0.850000\n",
      "L = 38.944961, acc = 0.850000\n",
      "L = 38.934453, acc = 0.850000\n",
      "L = 38.923993, acc = 0.855000\n",
      "L = 38.913579, acc = 0.855000\n",
      "L = 38.903210, acc = 0.855000\n",
      "L = 38.892883, acc = 0.855000\n",
      "L = 38.882595, acc = 0.855000\n",
      "L = 38.872344, acc = 0.855000\n",
      "L = 38.862129, acc = 0.855000\n",
      "L = 38.851946, acc = 0.855000\n",
      "L = 38.841794, acc = 0.855000\n",
      "L = 38.831671, acc = 0.855000\n",
      "L = 38.821574, acc = 0.855000\n",
      "L = 38.811503, acc = 0.855000\n",
      "L = 38.801454, acc = 0.855000\n",
      "L = 38.791426, acc = 0.855000\n",
      "L = 38.781418, acc = 0.855000\n",
      "L = 38.771427, acc = 0.855000\n",
      "L = 38.761452, acc = 0.855000\n",
      "L = 38.751491, acc = 0.855000\n",
      "L = 38.741542, acc = 0.855000\n",
      "L = 38.731604, acc = 0.855000\n",
      "L = 38.721676, acc = 0.855000\n",
      "L = 38.711755, acc = 0.855000\n",
      "L = 38.701840, acc = 0.855000\n",
      "L = 38.691929, acc = 0.855000\n",
      "L = 38.682022, acc = 0.855000\n",
      "L = 38.672117, acc = 0.855000\n",
      "L = 38.662212, acc = 0.855000\n",
      "L = 38.652306, acc = 0.855000\n",
      "L = 38.642397, acc = 0.855000\n",
      "L = 38.632485, acc = 0.855000\n",
      "L = 38.622568, acc = 0.855000\n",
      "L = 38.612645, acc = 0.855000\n",
      "L = 38.602715, acc = 0.855000\n",
      "L = 38.592775, acc = 0.855000\n",
      "L = 38.582826, acc = 0.855000\n",
      "L = 38.572866, acc = 0.855000\n",
      "L = 38.562894, acc = 0.855000\n",
      "L = 38.552908, acc = 0.855000\n",
      "L = 38.542908, acc = 0.855000\n",
      "L = 38.532892, acc = 0.855000\n",
      "L = 38.522860, acc = 0.855000\n",
      "L = 38.512811, acc = 0.855000\n",
      "L = 38.502742, acc = 0.855000\n",
      "L = 38.492655, acc = 0.855000\n",
      "L = 38.482546, acc = 0.855000\n",
      "L = 38.472416, acc = 0.855000\n",
      "L = 38.462263, acc = 0.855000\n",
      "L = 38.452087, acc = 0.855000\n",
      "L = 38.441886, acc = 0.855000\n",
      "L = 38.431660, acc = 0.855000\n",
      "L = 38.421407, acc = 0.855000\n",
      "L = 38.411128, acc = 0.855000\n",
      "L = 38.400820, acc = 0.855000\n",
      "L = 38.390483, acc = 0.855000\n",
      "L = 38.380116, acc = 0.855000\n",
      "L = 38.369719, acc = 0.855000\n",
      "L = 38.359290, acc = 0.855000\n",
      "L = 38.348829, acc = 0.855000\n",
      "L = 38.338334, acc = 0.855000\n",
      "L = 38.327806, acc = 0.855000\n",
      "L = 38.317242, acc = 0.855000\n",
      "L = 38.306643, acc = 0.855000\n",
      "L = 38.296008, acc = 0.855000\n",
      "L = 38.285335, acc = 0.855000\n",
      "L = 38.274625, acc = 0.855000\n",
      "L = 38.263875, acc = 0.855000\n",
      "L = 38.253086, acc = 0.855000\n",
      "L = 38.242257, acc = 0.855000\n",
      "L = 38.231387, acc = 0.855000\n",
      "L = 38.220475, acc = 0.855000\n",
      "L = 38.209520, acc = 0.855000\n",
      "L = 38.198523, acc = 0.855000\n",
      "L = 38.187481, acc = 0.855000\n",
      "L = 38.176394, acc = 0.855000\n",
      "L = 38.165262, acc = 0.855000\n",
      "L = 38.154084, acc = 0.855000\n",
      "L = 38.142859, acc = 0.855000\n",
      "L = 38.131586, acc = 0.855000\n",
      "L = 38.120265, acc = 0.855000\n",
      "L = 38.108895, acc = 0.855000\n",
      "L = 38.097475, acc = 0.855000\n",
      "L = 38.086004, acc = 0.855000\n",
      "L = 38.074483, acc = 0.855000\n",
      "L = 38.062909, acc = 0.855000\n",
      "L = 38.051283, acc = 0.855000\n",
      "L = 38.039603, acc = 0.855000\n",
      "L = 38.027870, acc = 0.855000\n",
      "L = 38.016082, acc = 0.855000\n",
      "L = 38.004238, acc = 0.855000\n",
      "L = 37.992338, acc = 0.860000\n",
      "L = 37.980381, acc = 0.860000\n",
      "L = 37.968367, acc = 0.860000\n",
      "L = 37.956295, acc = 0.860000\n",
      "L = 37.944163, acc = 0.860000\n",
      "L = 37.931972, acc = 0.860000\n",
      "L = 37.919720, acc = 0.860000\n",
      "L = 37.907408, acc = 0.860000\n",
      "L = 37.895033, acc = 0.860000\n",
      "L = 37.882596, acc = 0.860000\n",
      "L = 37.870096, acc = 0.860000\n",
      "L = 37.857532, acc = 0.860000\n",
      "L = 37.844903, acc = 0.860000\n",
      "L = 37.832209, acc = 0.860000\n",
      "L = 37.819448, acc = 0.860000\n",
      "L = 37.806621, acc = 0.860000\n",
      "L = 37.793727, acc = 0.860000\n",
      "L = 37.780763, acc = 0.860000\n",
      "L = 37.767731, acc = 0.860000\n",
      "L = 37.754630, acc = 0.860000\n",
      "L = 37.741457, acc = 0.860000\n",
      "L = 37.728214, acc = 0.865000\n",
      "L = 37.714898, acc = 0.865000\n",
      "L = 37.701510, acc = 0.865000\n",
      "L = 37.688048, acc = 0.865000\n",
      "L = 37.674512, acc = 0.865000\n",
      "L = 37.660902, acc = 0.865000\n",
      "L = 37.647215, acc = 0.865000\n",
      "L = 37.633452, acc = 0.865000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 37.619612, acc = 0.865000\n",
      "L = 37.605693, acc = 0.865000\n",
      "L = 37.591696, acc = 0.865000\n",
      "L = 37.577620, acc = 0.865000\n",
      "L = 37.563463, acc = 0.865000\n",
      "L = 37.549225, acc = 0.870000\n",
      "L = 37.534905, acc = 0.865000\n",
      "L = 37.520503, acc = 0.865000\n",
      "L = 37.506017, acc = 0.865000\n",
      "L = 37.491447, acc = 0.865000\n",
      "L = 37.476792, acc = 0.865000\n",
      "L = 37.462052, acc = 0.865000\n",
      "L = 37.447224, acc = 0.865000\n",
      "L = 37.432310, acc = 0.865000\n",
      "L = 37.417307, acc = 0.865000\n",
      "L = 37.402215, acc = 0.865000\n",
      "L = 37.387033, acc = 0.865000\n",
      "L = 37.371761, acc = 0.865000\n",
      "L = 37.356398, acc = 0.865000\n",
      "L = 37.340942, acc = 0.865000\n",
      "L = 37.325393, acc = 0.865000\n",
      "L = 37.309751, acc = 0.865000\n",
      "L = 37.294013, acc = 0.865000\n",
      "L = 37.278181, acc = 0.865000\n",
      "L = 37.262252, acc = 0.865000\n",
      "L = 37.246226, acc = 0.865000\n",
      "L = 37.230101, acc = 0.865000\n",
      "L = 37.213879, acc = 0.865000\n",
      "L = 37.197556, acc = 0.865000\n",
      "L = 37.181133, acc = 0.865000\n",
      "L = 37.164609, acc = 0.865000\n",
      "L = 37.147983, acc = 0.865000\n",
      "L = 37.131254, acc = 0.865000\n",
      "L = 37.114421, acc = 0.865000\n",
      "L = 37.097483, acc = 0.865000\n",
      "L = 37.080440, acc = 0.865000\n",
      "L = 37.063291, acc = 0.865000\n",
      "L = 37.046035, acc = 0.865000\n",
      "L = 37.028670, acc = 0.865000\n",
      "L = 37.011197, acc = 0.865000\n",
      "L = 36.993614, acc = 0.865000\n",
      "L = 36.975921, acc = 0.865000\n",
      "L = 36.958116, acc = 0.865000\n",
      "L = 36.940199, acc = 0.865000\n",
      "L = 36.922169, acc = 0.865000\n",
      "L = 36.904025, acc = 0.865000\n",
      "L = 36.885767, acc = 0.865000\n",
      "L = 36.867393, acc = 0.870000\n",
      "L = 36.848902, acc = 0.870000\n",
      "L = 36.830294, acc = 0.870000\n",
      "L = 36.811568, acc = 0.870000\n",
      "L = 36.792724, acc = 0.870000\n",
      "L = 36.773759, acc = 0.870000\n",
      "L = 36.754674, acc = 0.870000\n",
      "L = 36.735467, acc = 0.870000\n",
      "L = 36.716138, acc = 0.870000\n",
      "L = 36.696686, acc = 0.870000\n",
      "L = 36.677111, acc = 0.870000\n",
      "L = 36.657410, acc = 0.870000\n",
      "L = 36.637584, acc = 0.870000\n",
      "L = 36.617631, acc = 0.870000\n",
      "L = 36.597551, acc = 0.870000\n",
      "L = 36.577343, acc = 0.870000\n",
      "L = 36.557006, acc = 0.870000\n",
      "L = 36.536540, acc = 0.870000\n",
      "L = 36.515943, acc = 0.875000\n",
      "L = 36.495215, acc = 0.875000\n",
      "L = 36.474354, acc = 0.875000\n",
      "L = 36.453361, acc = 0.875000\n",
      "L = 36.432234, acc = 0.875000\n",
      "L = 36.410972, acc = 0.875000\n",
      "L = 36.389576, acc = 0.875000\n",
      "L = 36.368043, acc = 0.875000\n",
      "L = 36.346373, acc = 0.875000\n",
      "L = 36.324566, acc = 0.875000\n",
      "L = 36.302620, acc = 0.875000\n",
      "L = 36.280535, acc = 0.875000\n",
      "L = 36.258310, acc = 0.875000\n",
      "L = 36.235944, acc = 0.875000\n",
      "L = 36.213437, acc = 0.875000\n",
      "L = 36.190788, acc = 0.875000\n",
      "L = 36.167996, acc = 0.875000\n",
      "L = 36.145060, acc = 0.875000\n",
      "L = 36.121980, acc = 0.875000\n",
      "L = 36.098755, acc = 0.875000\n",
      "L = 36.075384, acc = 0.875000\n",
      "L = 36.051866, acc = 0.875000\n",
      "L = 36.028201, acc = 0.875000\n",
      "L = 36.004388, acc = 0.875000\n",
      "L = 35.980427, acc = 0.875000\n",
      "L = 35.956316, acc = 0.875000\n",
      "L = 35.932055, acc = 0.875000\n",
      "L = 35.907644, acc = 0.875000\n",
      "L = 35.883081, acc = 0.875000\n",
      "L = 35.858366, acc = 0.875000\n",
      "L = 35.833499, acc = 0.875000\n",
      "L = 35.808478, acc = 0.875000\n",
      "L = 35.783303, acc = 0.875000\n",
      "L = 35.757974, acc = 0.875000\n",
      "L = 35.732489, acc = 0.875000\n",
      "L = 35.706849, acc = 0.875000\n",
      "L = 35.681052, acc = 0.875000\n",
      "L = 35.655099, acc = 0.875000\n",
      "L = 35.628988, acc = 0.875000\n",
      "L = 35.602718, acc = 0.875000\n",
      "L = 35.576290, acc = 0.875000\n",
      "L = 35.549703, acc = 0.875000\n",
      "L = 35.522956, acc = 0.875000\n",
      "L = 35.496049, acc = 0.875000\n",
      "L = 35.468980, acc = 0.875000\n",
      "L = 35.441751, acc = 0.875000\n",
      "L = 35.414359, acc = 0.875000\n",
      "L = 35.386805, acc = 0.875000\n",
      "L = 35.359088, acc = 0.875000\n",
      "L = 35.331208, acc = 0.875000\n",
      "L = 35.303164, acc = 0.875000\n",
      "L = 35.274956, acc = 0.875000\n",
      "L = 35.246582, acc = 0.875000\n",
      "L = 35.218044, acc = 0.875000\n",
      "L = 35.189340, acc = 0.875000\n",
      "L = 35.160470, acc = 0.875000\n",
      "L = 35.131434, acc = 0.875000\n",
      "L = 35.102230, acc = 0.875000\n",
      "L = 35.072860, acc = 0.875000\n",
      "L = 35.043321, acc = 0.875000\n",
      "L = 35.013615, acc = 0.875000\n",
      "L = 34.983741, acc = 0.880000\n",
      "L = 34.953697, acc = 0.880000\n",
      "L = 34.923485, acc = 0.880000\n",
      "L = 34.893103, acc = 0.880000\n",
      "L = 34.862552, acc = 0.880000\n",
      "L = 34.831831, acc = 0.880000\n",
      "L = 34.800940, acc = 0.880000\n",
      "L = 34.769878, acc = 0.880000\n",
      "L = 34.738645, acc = 0.880000\n",
      "L = 34.707242, acc = 0.880000\n",
      "L = 34.675667, acc = 0.880000\n",
      "L = 34.643920, acc = 0.880000\n",
      "L = 34.612002, acc = 0.880000\n",
      "L = 34.579912, acc = 0.880000\n",
      "L = 34.547650, acc = 0.880000\n",
      "L = 34.515216, acc = 0.880000\n",
      "L = 34.482609, acc = 0.880000\n",
      "L = 34.449830, acc = 0.880000\n",
      "L = 34.416878, acc = 0.885000\n",
      "L = 34.383754, acc = 0.885000\n",
      "L = 34.350456, acc = 0.885000\n",
      "L = 34.316985, acc = 0.885000\n",
      "L = 34.283341, acc = 0.890000\n",
      "L = 34.249524, acc = 0.890000\n",
      "L = 34.215534, acc = 0.890000\n",
      "L = 34.181370, acc = 0.890000\n",
      "L = 34.147033, acc = 0.890000\n",
      "L = 34.112523, acc = 0.890000\n",
      "L = 34.077839, acc = 0.890000\n",
      "L = 34.042982, acc = 0.890000\n",
      "L = 34.007951, acc = 0.890000\n",
      "L = 33.972747, acc = 0.890000\n",
      "L = 33.937370, acc = 0.890000\n",
      "L = 33.901819, acc = 0.890000\n",
      "L = 33.866095, acc = 0.890000\n",
      "L = 33.830199, acc = 0.890000\n",
      "L = 33.794129, acc = 0.890000\n",
      "L = 33.757886, acc = 0.890000\n",
      "L = 33.721471, acc = 0.890000\n",
      "L = 33.684882, acc = 0.890000\n",
      "L = 33.648122, acc = 0.890000\n",
      "L = 33.611189, acc = 0.890000\n",
      "L = 33.574083, acc = 0.890000\n",
      "L = 33.536806, acc = 0.890000\n",
      "L = 33.499357, acc = 0.890000\n",
      "L = 33.461737, acc = 0.895000\n",
      "L = 33.423945, acc = 0.895000\n",
      "L = 33.385982, acc = 0.895000\n",
      "L = 33.347848, acc = 0.895000\n",
      "L = 33.309543, acc = 0.895000\n",
      "L = 33.271069, acc = 0.895000\n",
      "L = 33.232424, acc = 0.895000\n",
      "L = 33.193610, acc = 0.895000\n",
      "L = 33.154626, acc = 0.895000\n",
      "L = 33.115473, acc = 0.895000\n",
      "L = 33.076151, acc = 0.895000\n",
      "L = 33.036662, acc = 0.895000\n",
      "L = 32.997004, acc = 0.895000\n",
      "L = 32.957179, acc = 0.895000\n",
      "L = 32.917186, acc = 0.895000\n",
      "L = 32.877027, acc = 0.895000\n",
      "L = 32.836701, acc = 0.895000\n",
      "L = 32.796210, acc = 0.895000\n",
      "L = 32.755553, acc = 0.895000\n",
      "L = 32.714731, acc = 0.895000\n",
      "L = 32.673745, acc = 0.895000\n",
      "L = 32.632595, acc = 0.895000\n",
      "L = 32.591282, acc = 0.895000\n",
      "L = 32.549805, acc = 0.895000\n",
      "L = 32.508166, acc = 0.895000\n",
      "L = 32.466366, acc = 0.895000\n",
      "L = 32.424404, acc = 0.895000\n",
      "L = 32.382281, acc = 0.895000\n",
      "L = 32.339998, acc = 0.895000\n",
      "L = 32.297556, acc = 0.895000\n",
      "L = 32.254955, acc = 0.895000\n",
      "L = 32.212196, acc = 0.900000\n",
      "L = 32.169279, acc = 0.900000\n",
      "L = 32.126206, acc = 0.900000\n",
      "L = 32.082976, acc = 0.900000\n",
      "L = 32.039590, acc = 0.900000\n",
      "L = 31.996050, acc = 0.900000\n",
      "L = 31.952356, acc = 0.900000\n",
      "L = 31.908508, acc = 0.900000\n",
      "L = 31.864507, acc = 0.900000\n",
      "L = 31.820355, acc = 0.900000\n",
      "L = 31.776051, acc = 0.900000\n",
      "L = 31.731597, acc = 0.900000\n",
      "L = 31.686994, acc = 0.900000\n",
      "L = 31.642241, acc = 0.900000\n",
      "L = 31.597341, acc = 0.900000\n",
      "L = 31.552294, acc = 0.900000\n",
      "L = 31.507100, acc = 0.900000\n",
      "L = 31.461761, acc = 0.900000\n",
      "L = 31.416278, acc = 0.900000\n",
      "L = 31.370651, acc = 0.900000\n",
      "L = 31.324881, acc = 0.900000\n",
      "L = 31.278969, acc = 0.900000\n",
      "L = 31.232916, acc = 0.900000\n",
      "L = 31.186724, acc = 0.900000\n",
      "L = 31.140392, acc = 0.900000\n",
      "L = 31.093922, acc = 0.900000\n",
      "L = 31.047316, acc = 0.900000\n",
      "L = 31.000573, acc = 0.900000\n",
      "L = 30.953695, acc = 0.900000\n",
      "L = 30.906683, acc = 0.900000\n",
      "L = 30.859538, acc = 0.905000\n",
      "L = 30.812261, acc = 0.905000\n",
      "L = 30.764853, acc = 0.905000\n",
      "L = 30.717315, acc = 0.905000\n",
      "L = 30.669648, acc = 0.905000\n",
      "L = 30.621854, acc = 0.905000\n",
      "L = 30.573933, acc = 0.905000\n",
      "L = 30.525886, acc = 0.910000\n",
      "L = 30.477715, acc = 0.910000\n",
      "L = 30.429421, acc = 0.910000\n",
      "L = 30.381005, acc = 0.910000\n",
      "L = 30.332468, acc = 0.910000\n",
      "L = 30.283811, acc = 0.910000\n",
      "L = 30.235036, acc = 0.910000\n",
      "L = 30.186143, acc = 0.910000\n",
      "L = 30.137135, acc = 0.910000\n",
      "L = 30.088011, acc = 0.910000\n",
      "L = 30.038774, acc = 0.910000\n",
      "L = 29.989424, acc = 0.910000\n",
      "L = 29.939963, acc = 0.910000\n",
      "L = 29.890392, acc = 0.910000\n",
      "L = 29.840713, acc = 0.910000\n",
      "L = 29.790926, acc = 0.910000\n",
      "L = 29.741034, acc = 0.910000\n",
      "L = 29.691036, acc = 0.910000\n",
      "L = 29.640935, acc = 0.910000\n",
      "L = 29.590733, acc = 0.910000\n",
      "L = 29.540429, acc = 0.910000\n",
      "L = 29.490027, acc = 0.910000\n",
      "L = 29.439526, acc = 0.915000\n",
      "L = 29.388929, acc = 0.915000\n",
      "L = 29.338237, acc = 0.915000\n",
      "L = 29.287451, acc = 0.915000\n",
      "L = 29.236573, acc = 0.915000\n",
      "L = 29.185604, acc = 0.915000\n",
      "L = 29.134546, acc = 0.915000\n",
      "L = 29.083399, acc = 0.915000\n",
      "L = 29.032166, acc = 0.915000\n",
      "L = 28.980848, acc = 0.915000\n",
      "L = 28.929446, acc = 0.915000\n",
      "L = 28.877963, acc = 0.915000\n",
      "L = 28.826398, acc = 0.915000\n",
      "L = 28.774755, acc = 0.915000\n",
      "L = 28.723034, acc = 0.915000\n",
      "L = 28.671237, acc = 0.915000\n",
      "L = 28.619366, acc = 0.915000\n",
      "L = 28.567421, acc = 0.915000\n",
      "L = 28.515405, acc = 0.915000\n",
      "L = 28.463320, acc = 0.915000\n",
      "L = 28.411166, acc = 0.915000\n",
      "L = 28.358945, acc = 0.915000\n",
      "L = 28.306660, acc = 0.915000\n",
      "L = 28.254311, acc = 0.915000\n",
      "L = 28.201900, acc = 0.915000\n",
      "L = 28.149428, acc = 0.920000\n",
      "L = 28.096899, acc = 0.920000\n",
      "L = 28.044312, acc = 0.920000\n",
      "L = 27.991670, acc = 0.920000\n",
      "L = 27.938974, acc = 0.920000\n",
      "L = 27.886226, acc = 0.920000\n",
      "L = 27.833427, acc = 0.920000\n",
      "L = 27.780580, acc = 0.920000\n",
      "L = 27.727686, acc = 0.920000\n",
      "L = 27.674747, acc = 0.920000\n",
      "L = 27.621764, acc = 0.920000\n",
      "L = 27.568739, acc = 0.920000\n",
      "L = 27.515673, acc = 0.920000\n",
      "L = 27.462569, acc = 0.925000\n",
      "L = 27.409429, acc = 0.925000\n",
      "L = 27.356253, acc = 0.925000\n",
      "L = 27.303043, acc = 0.925000\n",
      "L = 27.249802, acc = 0.925000\n",
      "L = 27.196531, acc = 0.925000\n",
      "L = 27.143232, acc = 0.925000\n",
      "L = 27.089906, acc = 0.925000\n",
      "L = 27.036556, acc = 0.925000\n",
      "L = 26.983183, acc = 0.925000\n",
      "L = 26.929788, acc = 0.925000\n",
      "L = 26.876374, acc = 0.925000\n",
      "L = 26.822943, acc = 0.925000\n",
      "L = 26.769495, acc = 0.925000\n",
      "L = 26.716034, acc = 0.925000\n",
      "L = 26.662560, acc = 0.925000\n",
      "L = 26.609075, acc = 0.925000\n",
      "L = 26.555582, acc = 0.925000\n",
      "L = 26.502081, acc = 0.925000\n",
      "L = 26.448576, acc = 0.925000\n",
      "L = 26.395067, acc = 0.925000\n",
      "L = 26.341556, acc = 0.925000\n",
      "L = 26.288045, acc = 0.925000\n",
      "L = 26.234536, acc = 0.925000\n",
      "L = 26.181031, acc = 0.930000\n",
      "L = 26.127532, acc = 0.930000\n",
      "L = 26.074040, acc = 0.930000\n",
      "L = 26.020557, acc = 0.930000\n",
      "L = 25.967084, acc = 0.930000\n",
      "L = 25.913625, acc = 0.930000\n",
      "L = 25.860180, acc = 0.930000\n",
      "L = 25.806751, acc = 0.930000\n",
      "L = 25.753340, acc = 0.930000\n",
      "L = 25.699949, acc = 0.930000\n",
      "L = 25.646580, acc = 0.930000\n",
      "L = 25.593234, acc = 0.930000\n",
      "L = 25.539913, acc = 0.930000\n",
      "L = 25.486619, acc = 0.930000\n",
      "L = 25.433354, acc = 0.930000\n",
      "L = 25.380119, acc = 0.930000\n",
      "L = 25.326917, acc = 0.930000\n",
      "L = 25.273749, acc = 0.930000\n",
      "L = 25.220617, acc = 0.930000\n",
      "L = 25.167522, acc = 0.930000\n",
      "L = 25.114466, acc = 0.930000\n",
      "L = 25.061452, acc = 0.930000\n",
      "L = 25.008480, acc = 0.930000\n",
      "L = 24.955553, acc = 0.930000\n",
      "L = 24.902673, acc = 0.930000\n",
      "L = 24.849840, acc = 0.935000\n",
      "L = 24.797058, acc = 0.935000\n",
      "L = 24.744326, acc = 0.935000\n",
      "L = 24.691648, acc = 0.935000\n",
      "L = 24.639025, acc = 0.935000\n",
      "L = 24.586459, acc = 0.935000\n",
      "L = 24.533951, acc = 0.935000\n",
      "L = 24.481502, acc = 0.935000\n",
      "L = 24.429116, acc = 0.935000\n",
      "L = 24.376793, acc = 0.935000\n",
      "L = 24.324535, acc = 0.935000\n",
      "L = 24.272343, acc = 0.940000\n",
      "L = 24.220220, acc = 0.940000\n",
      "L = 24.168167, acc = 0.940000\n",
      "L = 24.116186, acc = 0.940000\n",
      "L = 24.064277, acc = 0.940000\n",
      "L = 24.012444, acc = 0.940000\n",
      "L = 23.960687, acc = 0.940000\n",
      "L = 23.909008, acc = 0.940000\n",
      "L = 23.857408, acc = 0.940000\n",
      "L = 23.805890, acc = 0.940000\n",
      "L = 23.754455, acc = 0.940000\n",
      "L = 23.703103, acc = 0.940000\n",
      "L = 23.651838, acc = 0.940000\n",
      "L = 23.600660, acc = 0.940000\n",
      "L = 23.549570, acc = 0.940000\n",
      "L = 23.498571, acc = 0.940000\n",
      "L = 23.447664, acc = 0.940000\n",
      "L = 23.396850, acc = 0.940000\n",
      "L = 23.346131, acc = 0.940000\n",
      "L = 23.295508, acc = 0.940000\n",
      "L = 23.244983, acc = 0.940000\n",
      "L = 23.194557, acc = 0.940000\n",
      "L = 23.144231, acc = 0.940000\n",
      "L = 23.094008, acc = 0.940000\n",
      "L = 23.043887, acc = 0.940000\n",
      "L = 22.993871, acc = 0.940000\n",
      "L = 22.943961, acc = 0.940000\n",
      "L = 22.894159, acc = 0.940000\n",
      "L = 22.844465, acc = 0.940000\n",
      "L = 22.794881, acc = 0.940000\n",
      "L = 22.745409, acc = 0.940000\n",
      "L = 22.696048, acc = 0.940000\n",
      "L = 22.646802, acc = 0.940000\n",
      "L = 22.597671, acc = 0.945000\n",
      "L = 22.548656, acc = 0.945000\n",
      "L = 22.499758, acc = 0.945000\n",
      "L = 22.450980, acc = 0.945000\n",
      "L = 22.402321, acc = 0.945000\n",
      "L = 22.353783, acc = 0.945000\n",
      "L = 22.305367, acc = 0.945000\n",
      "L = 22.257075, acc = 0.945000\n",
      "L = 22.208907, acc = 0.945000\n",
      "L = 22.160865, acc = 0.945000\n",
      "L = 22.112950, acc = 0.945000\n",
      "L = 22.065162, acc = 0.945000\n",
      "L = 22.017503, acc = 0.945000\n",
      "L = 21.969973, acc = 0.945000\n",
      "L = 21.922575, acc = 0.945000\n",
      "L = 21.875308, acc = 0.945000\n",
      "L = 21.828175, acc = 0.945000\n",
      "L = 21.781174, acc = 0.945000\n",
      "L = 21.734309, acc = 0.945000\n",
      "L = 21.687579, acc = 0.945000\n",
      "L = 21.640986, acc = 0.945000\n",
      "L = 21.594530, acc = 0.945000\n",
      "L = 21.548213, acc = 0.945000\n",
      "L = 21.502034, acc = 0.945000\n",
      "L = 21.455996, acc = 0.945000\n",
      "L = 21.410098, acc = 0.945000\n",
      "L = 21.364342, acc = 0.945000\n",
      "L = 21.318729, acc = 0.945000\n",
      "L = 21.273258, acc = 0.945000\n",
      "L = 21.227932, acc = 0.945000\n",
      "L = 21.182750, acc = 0.945000\n",
      "L = 21.137714, acc = 0.945000\n",
      "L = 21.092823, acc = 0.945000\n",
      "L = 21.048079, acc = 0.945000\n",
      "L = 21.003483, acc = 0.945000\n",
      "L = 20.959034, acc = 0.945000\n",
      "L = 20.914734, acc = 0.945000\n",
      "L = 20.870584, acc = 0.945000\n",
      "L = 20.826583, acc = 0.945000\n",
      "L = 20.782732, acc = 0.945000\n",
      "L = 20.739032, acc = 0.945000\n",
      "L = 20.695484, acc = 0.945000\n",
      "L = 20.652088, acc = 0.945000\n",
      "L = 20.608844, acc = 0.945000\n",
      "L = 20.565752, acc = 0.945000\n",
      "L = 20.522815, acc = 0.945000\n",
      "L = 20.480031, acc = 0.945000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 20.437401, acc = 0.945000\n",
      "L = 20.394926, acc = 0.945000\n",
      "L = 20.352605, acc = 0.945000\n",
      "L = 20.310440, acc = 0.945000\n",
      "L = 20.268431, acc = 0.945000\n",
      "L = 20.226578, acc = 0.945000\n",
      "L = 20.184881, acc = 0.945000\n",
      "L = 20.143341, acc = 0.945000\n",
      "L = 20.101957, acc = 0.945000\n",
      "L = 20.060731, acc = 0.945000\n",
      "L = 20.019662, acc = 0.945000\n",
      "L = 19.978751, acc = 0.945000\n",
      "L = 19.937997, acc = 0.945000\n",
      "L = 19.897402, acc = 0.945000\n",
      "L = 19.856965, acc = 0.945000\n",
      "L = 19.816686, acc = 0.945000\n",
      "L = 19.776566, acc = 0.945000\n",
      "L = 19.736604, acc = 0.945000\n",
      "L = 19.696802, acc = 0.945000\n",
      "L = 19.657158, acc = 0.945000\n",
      "L = 19.617673, acc = 0.945000\n",
      "L = 19.578347, acc = 0.945000\n",
      "L = 19.539180, acc = 0.945000\n",
      "L = 19.500173, acc = 0.945000\n",
      "L = 19.461324, acc = 0.945000\n",
      "L = 19.422635, acc = 0.945000\n",
      "L = 19.384105, acc = 0.945000\n",
      "L = 19.345734, acc = 0.945000\n",
      "L = 19.307522, acc = 0.945000\n",
      "L = 19.269469, acc = 0.945000\n",
      "L = 19.231575, acc = 0.945000\n",
      "L = 19.193840, acc = 0.945000\n",
      "L = 19.156264, acc = 0.945000\n",
      "L = 19.118847, acc = 0.945000\n",
      "L = 19.081588, acc = 0.945000\n",
      "L = 19.044488, acc = 0.945000\n",
      "L = 19.007546, acc = 0.945000\n",
      "L = 18.970763, acc = 0.945000\n",
      "L = 18.934137, acc = 0.945000\n",
      "L = 18.897670, acc = 0.945000\n",
      "L = 18.861360, acc = 0.945000\n",
      "L = 18.825208, acc = 0.945000\n",
      "L = 18.789212, acc = 0.945000\n",
      "L = 18.753374, acc = 0.945000\n",
      "L = 18.717693, acc = 0.945000\n",
      "L = 18.682169, acc = 0.945000\n",
      "L = 18.646800, acc = 0.945000\n",
      "L = 18.611588, acc = 0.945000\n",
      "L = 18.576532, acc = 0.945000\n",
      "L = 18.541631, acc = 0.945000\n",
      "L = 18.506885, acc = 0.945000\n",
      "L = 18.472295, acc = 0.945000\n",
      "L = 18.437858, acc = 0.945000\n",
      "L = 18.403577, acc = 0.945000\n",
      "L = 18.369449, acc = 0.945000\n",
      "L = 18.335475, acc = 0.945000\n",
      "L = 18.301654, acc = 0.945000\n",
      "L = 18.267985, acc = 0.945000\n",
      "L = 18.234470, acc = 0.945000\n",
      "L = 18.201106, acc = 0.945000\n",
      "L = 18.167895, acc = 0.945000\n",
      "L = 18.134835, acc = 0.945000\n",
      "L = 18.101925, acc = 0.945000\n",
      "L = 18.069167, acc = 0.945000\n",
      "L = 18.036558, acc = 0.945000\n",
      "L = 18.004099, acc = 0.945000\n",
      "L = 17.971790, acc = 0.945000\n",
      "L = 17.939629, acc = 0.945000\n",
      "L = 17.907617, acc = 0.945000\n",
      "L = 17.875753, acc = 0.945000\n",
      "L = 17.844036, acc = 0.945000\n",
      "L = 17.812467, acc = 0.945000\n",
      "L = 17.781044, acc = 0.945000\n",
      "L = 17.749767, acc = 0.945000\n",
      "L = 17.718636, acc = 0.945000\n",
      "L = 17.687650, acc = 0.945000\n",
      "L = 17.656809, acc = 0.945000\n",
      "L = 17.626111, acc = 0.945000\n",
      "L = 17.595558, acc = 0.945000\n",
      "L = 17.565148, acc = 0.945000\n",
      "L = 17.534880, acc = 0.945000\n",
      "L = 17.504755, acc = 0.945000\n",
      "L = 17.474771, acc = 0.945000\n",
      "L = 17.444929, acc = 0.945000\n",
      "L = 17.415227, acc = 0.945000\n",
      "L = 17.385665, acc = 0.945000\n",
      "L = 17.356243, acc = 0.945000\n",
      "L = 17.326960, acc = 0.945000\n",
      "L = 17.297815, acc = 0.945000\n",
      "L = 17.268808, acc = 0.945000\n",
      "L = 17.239939, acc = 0.945000\n",
      "L = 17.211206, acc = 0.945000\n",
      "L = 17.182610, acc = 0.950000\n",
      "L = 17.154149, acc = 0.950000\n",
      "L = 17.125824, acc = 0.950000\n",
      "L = 17.097633, acc = 0.950000\n",
      "L = 17.069577, acc = 0.950000\n",
      "L = 17.041653, acc = 0.950000\n",
      "L = 17.013863, acc = 0.950000\n",
      "L = 16.986205, acc = 0.950000\n",
      "L = 16.958679, acc = 0.950000\n",
      "L = 16.931284, acc = 0.950000\n",
      "L = 16.904020, acc = 0.950000\n",
      "L = 16.876886, acc = 0.950000\n",
      "L = 16.849881, acc = 0.950000\n",
      "L = 16.823006, acc = 0.950000\n",
      "L = 16.796258, acc = 0.950000\n",
      "L = 16.769639, acc = 0.950000\n",
      "L = 16.743146, acc = 0.950000\n",
      "L = 16.716780, acc = 0.950000\n",
      "L = 16.690540, acc = 0.950000\n",
      "L = 16.664426, acc = 0.950000\n",
      "L = 16.638436, acc = 0.950000\n",
      "L = 16.612571, acc = 0.950000\n",
      "L = 16.586829, acc = 0.950000\n",
      "L = 16.561211, acc = 0.950000\n",
      "L = 16.535715, acc = 0.950000\n",
      "L = 16.510341, acc = 0.950000\n",
      "L = 16.485088, acc = 0.950000\n",
      "L = 16.459956, acc = 0.950000\n",
      "L = 16.434944, acc = 0.950000\n",
      "L = 16.410051, acc = 0.950000\n",
      "L = 16.385278, acc = 0.950000\n",
      "L = 16.360623, acc = 0.950000\n",
      "L = 16.336085, acc = 0.950000\n",
      "L = 16.311665, acc = 0.950000\n",
      "L = 16.287362, acc = 0.950000\n",
      "L = 16.263175, acc = 0.950000\n",
      "L = 16.239103, acc = 0.955000\n",
      "L = 16.215146, acc = 0.955000\n",
      "L = 16.191303, acc = 0.955000\n",
      "L = 16.167574, acc = 0.955000\n",
      "L = 16.143958, acc = 0.955000\n",
      "L = 16.120455, acc = 0.955000\n",
      "L = 16.097064, acc = 0.955000\n",
      "L = 16.073784, acc = 0.955000\n",
      "L = 16.050615, acc = 0.955000\n",
      "L = 16.027556, acc = 0.955000\n",
      "L = 16.004606, acc = 0.955000\n",
      "L = 15.981766, acc = 0.955000\n",
      "L = 15.959035, acc = 0.955000\n",
      "L = 15.936411, acc = 0.955000\n",
      "L = 15.913895, acc = 0.955000\n",
      "L = 15.891485, acc = 0.955000\n",
      "L = 15.869182, acc = 0.955000\n",
      "L = 15.846984, acc = 0.955000\n",
      "L = 15.824892, acc = 0.955000\n",
      "L = 15.802904, acc = 0.955000\n",
      "L = 15.781020, acc = 0.955000\n",
      "L = 15.759239, acc = 0.955000\n",
      "L = 15.737561, acc = 0.955000\n",
      "L = 15.715986, acc = 0.955000\n",
      "L = 15.694512, acc = 0.955000\n",
      "L = 15.673140, acc = 0.955000\n",
      "L = 15.651868, acc = 0.955000\n",
      "L = 15.630696, acc = 0.955000\n",
      "L = 15.609624, acc = 0.955000\n",
      "L = 15.588651, acc = 0.955000\n",
      "L = 15.567776, acc = 0.955000\n",
      "L = 15.546999, acc = 0.955000\n",
      "L = 15.526320, acc = 0.955000\n",
      "L = 15.505737, acc = 0.955000\n",
      "L = 15.485251, acc = 0.955000\n",
      "L = 15.464860, acc = 0.955000\n",
      "L = 15.444565, acc = 0.955000\n",
      "L = 15.424364, acc = 0.955000\n",
      "L = 15.404258, acc = 0.955000\n",
      "L = 15.384245, acc = 0.955000\n",
      "L = 15.364325, acc = 0.955000\n",
      "L = 15.344498, acc = 0.955000\n",
      "L = 15.324763, acc = 0.955000\n",
      "L = 15.305119, acc = 0.955000\n",
      "L = 15.285567, acc = 0.955000\n",
      "L = 15.266105, acc = 0.955000\n",
      "L = 15.246733, acc = 0.955000\n",
      "L = 15.227450, acc = 0.955000\n",
      "L = 15.208257, acc = 0.955000\n",
      "L = 15.189152, acc = 0.955000\n",
      "L = 15.170135, acc = 0.955000\n",
      "L = 15.151205, acc = 0.955000\n",
      "L = 15.132362, acc = 0.955000\n",
      "L = 15.113606, acc = 0.955000\n",
      "L = 15.094936, acc = 0.955000\n",
      "L = 15.076352, acc = 0.955000\n",
      "L = 15.057852, acc = 0.955000\n",
      "L = 15.039437, acc = 0.955000\n",
      "L = 15.021106, acc = 0.955000\n",
      "L = 15.002859, acc = 0.955000\n",
      "L = 14.984695, acc = 0.955000\n",
      "L = 14.966613, acc = 0.955000\n",
      "L = 14.948613, acc = 0.960000\n",
      "L = 14.930695, acc = 0.960000\n",
      "L = 14.912859, acc = 0.960000\n",
      "L = 14.895103, acc = 0.960000\n",
      "L = 14.877427, acc = 0.960000\n",
      "L = 14.859831, acc = 0.960000\n",
      "L = 14.842314, acc = 0.960000\n",
      "L = 14.824876, acc = 0.960000\n",
      "L = 14.807517, acc = 0.960000\n",
      "L = 14.790236, acc = 0.960000\n",
      "L = 14.773032, acc = 0.960000\n",
      "L = 14.755905, acc = 0.960000\n",
      "L = 14.738855, acc = 0.960000\n",
      "L = 14.721881, acc = 0.960000\n",
      "L = 14.704982, acc = 0.960000\n",
      "L = 14.688159, acc = 0.960000\n",
      "L = 14.671411, acc = 0.960000\n",
      "L = 14.654737, acc = 0.960000\n",
      "L = 14.638137, acc = 0.960000\n",
      "L = 14.621611, acc = 0.960000\n",
      "L = 14.605157, acc = 0.960000\n",
      "L = 14.588777, acc = 0.960000\n",
      "L = 14.572468, acc = 0.960000\n",
      "L = 14.556232, acc = 0.960000\n",
      "L = 14.540067, acc = 0.960000\n",
      "L = 14.523973, acc = 0.960000\n",
      "L = 14.507949, acc = 0.960000\n",
      "L = 14.491996, acc = 0.960000\n",
      "L = 14.476113, acc = 0.960000\n",
      "L = 14.460298, acc = 0.960000\n",
      "L = 14.444553, acc = 0.960000\n",
      "L = 14.428877, acc = 0.960000\n",
      "L = 14.413268, acc = 0.960000\n",
      "L = 14.397727, acc = 0.960000\n",
      "L = 14.382254, acc = 0.960000\n",
      "L = 14.366847, acc = 0.960000\n",
      "L = 14.351507, acc = 0.960000\n",
      "L = 14.336234, acc = 0.960000\n",
      "L = 14.321026, acc = 0.960000\n",
      "L = 14.305883, acc = 0.960000\n",
      "L = 14.290805, acc = 0.960000\n",
      "L = 14.275793, acc = 0.960000\n",
      "L = 14.260844, acc = 0.960000\n",
      "L = 14.245959, acc = 0.960000\n",
      "L = 14.231138, acc = 0.960000\n",
      "L = 14.216380, acc = 0.960000\n",
      "L = 14.201684, acc = 0.960000\n",
      "L = 14.187051, acc = 0.960000\n",
      "L = 14.172480, acc = 0.960000\n",
      "L = 14.157971, acc = 0.960000\n",
      "L = 14.143523, acc = 0.960000\n",
      "L = 14.129136, acc = 0.960000\n",
      "L = 14.114810, acc = 0.960000\n",
      "L = 14.100543, acc = 0.960000\n",
      "L = 14.086337, acc = 0.960000\n",
      "L = 14.072190, acc = 0.960000\n",
      "L = 14.058102, acc = 0.960000\n",
      "L = 14.044074, acc = 0.960000\n",
      "L = 14.030103, acc = 0.960000\n",
      "L = 14.016191, acc = 0.960000\n",
      "L = 14.002337, acc = 0.960000\n",
      "L = 13.988540, acc = 0.960000\n",
      "L = 13.974800, acc = 0.960000\n",
      "L = 13.961117, acc = 0.960000\n",
      "L = 13.947491, acc = 0.960000\n",
      "L = 13.933920, acc = 0.960000\n",
      "L = 13.920406, acc = 0.960000\n",
      "L = 13.906947, acc = 0.960000\n",
      "L = 13.893543, acc = 0.960000\n",
      "L = 13.880194, acc = 0.960000\n",
      "L = 13.866899, acc = 0.960000\n",
      "L = 13.853659, acc = 0.960000\n",
      "L = 13.840472, acc = 0.960000\n",
      "L = 13.827339, acc = 0.960000\n",
      "L = 13.814260, acc = 0.960000\n",
      "L = 13.801233, acc = 0.960000\n",
      "L = 13.788259, acc = 0.960000\n",
      "L = 13.775337, acc = 0.960000\n",
      "L = 13.762467, acc = 0.965000\n",
      "L = 13.749649, acc = 0.965000\n",
      "L = 13.736883, acc = 0.965000\n",
      "L = 13.724167, acc = 0.965000\n",
      "L = 13.711503, acc = 0.965000\n",
      "L = 13.698888, acc = 0.965000\n",
      "L = 13.686324, acc = 0.965000\n",
      "L = 13.673810, acc = 0.965000\n",
      "L = 13.661346, acc = 0.965000\n",
      "L = 13.648931, acc = 0.965000\n",
      "L = 13.636565, acc = 0.965000\n",
      "L = 13.624248, acc = 0.965000\n",
      "L = 13.611980, acc = 0.965000\n",
      "L = 13.599759, acc = 0.965000\n",
      "L = 13.587587, acc = 0.965000\n",
      "L = 13.575462, acc = 0.965000\n",
      "L = 13.563385, acc = 0.965000\n",
      "L = 13.551354, acc = 0.965000\n",
      "L = 13.539371, acc = 0.965000\n",
      "L = 13.527434, acc = 0.965000\n",
      "L = 13.515544, acc = 0.965000\n",
      "L = 13.503699, acc = 0.965000\n",
      "L = 13.491901, acc = 0.965000\n",
      "L = 13.480147, acc = 0.965000\n",
      "L = 13.468439, acc = 0.965000\n",
      "L = 13.456776, acc = 0.965000\n",
      "L = 13.445158, acc = 0.965000\n",
      "L = 13.433585, acc = 0.965000\n",
      "L = 13.422055, acc = 0.965000\n",
      "L = 13.410569, acc = 0.965000\n",
      "L = 13.399128, acc = 0.965000\n",
      "L = 13.387729, acc = 0.965000\n",
      "L = 13.376374, acc = 0.965000\n",
      "L = 13.365062, acc = 0.965000\n",
      "L = 13.353792, acc = 0.965000\n",
      "L = 13.342565, acc = 0.965000\n",
      "L = 13.331380, acc = 0.965000\n",
      "L = 13.320237, acc = 0.965000\n",
      "L = 13.309136, acc = 0.965000\n",
      "L = 13.298077, acc = 0.965000\n",
      "L = 13.287059, acc = 0.965000\n",
      "L = 13.276081, acc = 0.965000\n",
      "L = 13.265145, acc = 0.965000\n",
      "L = 13.254249, acc = 0.965000\n",
      "L = 13.243393, acc = 0.965000\n",
      "L = 13.232578, acc = 0.965000\n",
      "L = 13.221802, acc = 0.965000\n",
      "L = 13.211066, acc = 0.965000\n",
      "L = 13.200370, acc = 0.965000\n",
      "L = 13.189712, acc = 0.965000\n",
      "L = 13.179094, acc = 0.965000\n",
      "L = 13.168515, acc = 0.965000\n",
      "L = 13.157973, acc = 0.965000\n",
      "L = 13.147471, acc = 0.965000\n",
      "L = 13.137006, acc = 0.965000\n",
      "L = 13.126579, acc = 0.965000\n",
      "L = 13.116190, acc = 0.965000\n",
      "L = 13.105839, acc = 0.965000\n",
      "L = 13.095525, acc = 0.965000\n",
      "L = 13.085247, acc = 0.965000\n",
      "L = 13.075007, acc = 0.965000\n",
      "L = 13.064803, acc = 0.965000\n",
      "L = 13.054636, acc = 0.965000\n",
      "L = 13.044504, acc = 0.965000\n",
      "L = 13.034409, acc = 0.965000\n",
      "L = 13.024350, acc = 0.965000\n",
      "L = 13.014326, acc = 0.965000\n",
      "L = 13.004338, acc = 0.965000\n",
      "L = 12.994385, acc = 0.965000\n",
      "L = 12.984467, acc = 0.965000\n",
      "L = 12.974583, acc = 0.965000\n",
      "L = 12.964735, acc = 0.965000\n",
      "L = 12.954920, acc = 0.965000\n",
      "L = 12.945140, acc = 0.965000\n",
      "L = 12.935394, acc = 0.965000\n",
      "L = 12.925682, acc = 0.965000\n",
      "L = 12.916004, acc = 0.965000\n",
      "L = 12.906359, acc = 0.965000\n",
      "L = 12.896747, acc = 0.965000\n",
      "L = 12.887169, acc = 0.965000\n",
      "L = 12.877623, acc = 0.965000\n",
      "L = 12.868110, acc = 0.965000\n",
      "L = 12.858630, acc = 0.965000\n",
      "L = 12.849182, acc = 0.965000\n",
      "L = 12.839766, acc = 0.965000\n",
      "L = 12.830383, acc = 0.965000\n",
      "L = 12.821031, acc = 0.965000\n",
      "L = 12.811711, acc = 0.965000\n",
      "L = 12.802422, acc = 0.965000\n",
      "L = 12.793165, acc = 0.965000\n",
      "L = 12.783939, acc = 0.965000\n",
      "L = 12.774743, acc = 0.965000\n",
      "L = 12.765579, acc = 0.965000\n",
      "L = 12.756445, acc = 0.965000\n",
      "L = 12.747342, acc = 0.965000\n",
      "L = 12.738269, acc = 0.965000\n",
      "L = 12.729226, acc = 0.965000\n",
      "L = 12.720214, acc = 0.965000\n",
      "L = 12.711231, acc = 0.965000\n",
      "L = 12.702277, acc = 0.965000\n",
      "L = 12.693354, acc = 0.965000\n",
      "L = 12.684459, acc = 0.965000\n",
      "L = 12.675594, acc = 0.965000\n",
      "L = 12.666757, acc = 0.965000\n",
      "L = 12.657950, acc = 0.965000\n",
      "L = 12.649171, acc = 0.965000\n",
      "L = 12.640421, acc = 0.965000\n",
      "L = 12.631699, acc = 0.965000\n",
      "L = 12.623006, acc = 0.965000\n",
      "L = 12.614340, acc = 0.965000\n",
      "L = 12.605703, acc = 0.965000\n",
      "L = 12.597093, acc = 0.965000\n",
      "L = 12.588511, acc = 0.965000\n",
      "L = 12.579956, acc = 0.965000\n",
      "L = 12.571429, acc = 0.965000\n",
      "L = 12.562928, acc = 0.965000\n",
      "L = 12.554455, acc = 0.965000\n",
      "L = 12.546009, acc = 0.965000\n",
      "L = 12.537590, acc = 0.965000\n",
      "L = 12.529197, acc = 0.965000\n",
      "L = 12.520831, acc = 0.965000\n",
      "L = 12.512491, acc = 0.965000\n",
      "L = 12.504177, acc = 0.965000\n",
      "L = 12.495889, acc = 0.965000\n",
      "L = 12.487627, acc = 0.965000\n",
      "L = 12.479391, acc = 0.965000\n",
      "L = 12.471180, acc = 0.965000\n",
      "L = 12.462995, acc = 0.965000\n",
      "L = 12.454836, acc = 0.965000\n",
      "L = 12.446701, acc = 0.965000\n",
      "L = 12.438592, acc = 0.965000\n",
      "L = 12.430508, acc = 0.965000\n",
      "L = 12.422448, acc = 0.965000\n",
      "L = 12.414413, acc = 0.965000\n",
      "L = 12.406403, acc = 0.965000\n",
      "L = 12.398417, acc = 0.965000\n",
      "L = 12.390456, acc = 0.965000\n",
      "L = 12.382519, acc = 0.965000\n",
      "L = 12.374605, acc = 0.965000\n",
      "L = 12.366716, acc = 0.965000\n",
      "L = 12.358851, acc = 0.965000\n",
      "L = 12.351009, acc = 0.965000\n",
      "L = 12.343190, acc = 0.965000\n",
      "L = 12.335396, acc = 0.965000\n",
      "L = 12.327624, acc = 0.965000\n",
      "L = 12.319876, acc = 0.965000\n",
      "L = 12.312151, acc = 0.965000\n",
      "L = 12.304448, acc = 0.965000\n",
      "L = 12.296769, acc = 0.965000\n",
      "L = 12.289112, acc = 0.965000\n",
      "L = 12.281478, acc = 0.965000\n",
      "L = 12.273866, acc = 0.965000\n",
      "L = 12.266277, acc = 0.965000\n",
      "L = 12.258710, acc = 0.965000\n",
      "L = 12.251165, acc = 0.965000\n",
      "L = 12.243642, acc = 0.965000\n",
      "L = 12.236141, acc = 0.965000\n",
      "L = 12.228662, acc = 0.965000\n",
      "L = 12.221204, acc = 0.965000\n",
      "L = 12.213768, acc = 0.965000\n",
      "L = 12.206354, acc = 0.965000\n",
      "L = 12.198961, acc = 0.965000\n",
      "L = 12.191589, acc = 0.965000\n",
      "L = 12.184238, acc = 0.965000\n",
      "L = 12.176908, acc = 0.965000\n",
      "L = 12.169599, acc = 0.965000\n",
      "L = 12.162311, acc = 0.965000\n",
      "L = 12.155044, acc = 0.965000\n",
      "L = 12.147797, acc = 0.965000\n",
      "L = 12.140571, acc = 0.965000\n",
      "L = 12.133365, acc = 0.965000\n",
      "L = 12.126180, acc = 0.965000\n",
      "L = 12.119015, acc = 0.965000\n",
      "L = 12.111869, acc = 0.965000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 12.104744, acc = 0.965000\n",
      "L = 12.097639, acc = 0.965000\n",
      "L = 12.090553, acc = 0.965000\n",
      "L = 12.083487, acc = 0.965000\n",
      "L = 12.076441, acc = 0.965000\n",
      "L = 12.069414, acc = 0.965000\n",
      "L = 12.062407, acc = 0.965000\n",
      "L = 12.055419, acc = 0.965000\n",
      "L = 12.048450, acc = 0.965000\n",
      "L = 12.041500, acc = 0.965000\n",
      "L = 12.034569, acc = 0.965000\n",
      "L = 12.027657, acc = 0.965000\n",
      "L = 12.020764, acc = 0.965000\n",
      "L = 12.013890, acc = 0.965000\n",
      "L = 12.007034, acc = 0.965000\n",
      "L = 12.000197, acc = 0.965000\n",
      "L = 11.993379, acc = 0.965000\n",
      "L = 11.986578, acc = 0.965000\n",
      "L = 11.979796, acc = 0.965000\n",
      "L = 11.973033, acc = 0.965000\n",
      "L = 11.966287, acc = 0.965000\n",
      "L = 11.959559, acc = 0.965000\n",
      "L = 11.952849, acc = 0.965000\n",
      "L = 11.946157, acc = 0.965000\n",
      "L = 11.939483, acc = 0.965000\n",
      "L = 11.932827, acc = 0.965000\n",
      "L = 11.926188, acc = 0.965000\n",
      "L = 11.919566, acc = 0.965000\n",
      "L = 11.912962, acc = 0.965000\n",
      "L = 11.906376, acc = 0.965000\n",
      "L = 11.899806, acc = 0.965000\n",
      "L = 11.893254, acc = 0.965000\n",
      "L = 11.886718, acc = 0.965000\n",
      "L = 11.880200, acc = 0.965000\n",
      "L = 11.873699, acc = 0.965000\n",
      "L = 11.867214, acc = 0.965000\n",
      "L = 11.860747, acc = 0.965000\n",
      "L = 11.854295, acc = 0.965000\n",
      "L = 11.847861, acc = 0.965000\n",
      "L = 11.841443, acc = 0.965000\n",
      "L = 11.835041, acc = 0.965000\n",
      "L = 11.828656, acc = 0.965000\n",
      "L = 11.822287, acc = 0.965000\n",
      "L = 11.815935, acc = 0.965000\n",
      "L = 11.809598, acc = 0.965000\n",
      "L = 11.803278, acc = 0.965000\n",
      "L = 11.796973, acc = 0.965000\n",
      "L = 11.790684, acc = 0.965000\n",
      "L = 11.784412, acc = 0.965000\n",
      "L = 11.778154, acc = 0.965000\n",
      "L = 11.771913, acc = 0.965000\n",
      "L = 11.765687, acc = 0.965000\n",
      "L = 11.759477, acc = 0.965000\n",
      "L = 11.753282, acc = 0.970000\n",
      "L = 11.747103, acc = 0.970000\n",
      "L = 11.740939, acc = 0.970000\n",
      "L = 11.734790, acc = 0.970000\n",
      "L = 11.728656, acc = 0.970000\n",
      "L = 11.722538, acc = 0.970000\n",
      "L = 11.716434, acc = 0.970000\n",
      "L = 11.710346, acc = 0.970000\n",
      "L = 11.704272, acc = 0.970000\n",
      "L = 11.698213, acc = 0.970000\n",
      "L = 11.692169, acc = 0.970000\n",
      "L = 11.686140, acc = 0.970000\n",
      "L = 11.680125, acc = 0.970000\n",
      "L = 11.674125, acc = 0.970000\n",
      "L = 11.668140, acc = 0.970000\n",
      "L = 11.662169, acc = 0.970000\n",
      "L = 11.656212, acc = 0.970000\n",
      "L = 11.650269, acc = 0.970000\n",
      "L = 11.644341, acc = 0.970000\n",
      "L = 11.638427, acc = 0.970000\n",
      "L = 11.632527, acc = 0.970000\n",
      "L = 11.626641, acc = 0.970000\n",
      "L = 11.620769, acc = 0.970000\n",
      "L = 11.614911, acc = 0.970000\n",
      "L = 11.609067, acc = 0.970000\n",
      "L = 11.603237, acc = 0.970000\n",
      "L = 11.597420, acc = 0.970000\n",
      "L = 11.591618, acc = 0.970000\n",
      "L = 11.585828, acc = 0.970000\n",
      "L = 11.580053, acc = 0.970000\n",
      "L = 11.574291, acc = 0.970000\n",
      "L = 11.568542, acc = 0.970000\n",
      "L = 11.562807, acc = 0.970000\n",
      "L = 11.557085, acc = 0.970000\n",
      "L = 11.551376, acc = 0.970000\n",
      "L = 11.545680, acc = 0.970000\n",
      "L = 11.539998, acc = 0.970000\n",
      "L = 11.534329, acc = 0.970000\n",
      "L = 11.528673, acc = 0.970000\n",
      "L = 11.523029, acc = 0.970000\n",
      "L = 11.517399, acc = 0.970000\n",
      "L = 11.511782, acc = 0.970000\n",
      "L = 11.506177, acc = 0.970000\n",
      "L = 11.500585, acc = 0.970000\n",
      "L = 11.495006, acc = 0.970000\n",
      "L = 11.489440, acc = 0.970000\n",
      "L = 11.483886, acc = 0.970000\n",
      "L = 11.478345, acc = 0.970000\n",
      "L = 11.472816, acc = 0.970000\n",
      "L = 11.467300, acc = 0.970000\n",
      "L = 11.461796, acc = 0.970000\n",
      "L = 11.456304, acc = 0.970000\n",
      "L = 11.450825, acc = 0.970000\n",
      "L = 11.445358, acc = 0.970000\n",
      "L = 11.439903, acc = 0.970000\n",
      "L = 11.434461, acc = 0.970000\n",
      "L = 11.429030, acc = 0.970000\n",
      "L = 11.423612, acc = 0.970000\n",
      "L = 11.418205, acc = 0.970000\n",
      "L = 11.412810, acc = 0.970000\n",
      "L = 11.407428, acc = 0.970000\n",
      "L = 11.402057, acc = 0.970000\n",
      "L = 11.396698, acc = 0.970000\n",
      "L = 11.391351, acc = 0.970000\n",
      "L = 11.386015, acc = 0.970000\n",
      "L = 11.380691, acc = 0.970000\n",
      "L = 11.375379, acc = 0.970000\n",
      "L = 11.370078, acc = 0.970000\n",
      "L = 11.364789, acc = 0.970000\n",
      "L = 11.359511, acc = 0.970000\n",
      "L = 11.354245, acc = 0.970000\n",
      "L = 11.348990, acc = 0.970000\n",
      "L = 11.343746, acc = 0.970000\n",
      "L = 11.338514, acc = 0.970000\n",
      "L = 11.333293, acc = 0.970000\n",
      "L = 11.328083, acc = 0.970000\n",
      "L = 11.322884, acc = 0.970000\n",
      "L = 11.317696, acc = 0.970000\n",
      "L = 11.312520, acc = 0.970000\n",
      "L = 11.307354, acc = 0.970000\n",
      "L = 11.302200, acc = 0.970000\n",
      "L = 11.297056, acc = 0.970000\n",
      "L = 11.291923, acc = 0.970000\n",
      "L = 11.286802, acc = 0.970000\n",
      "L = 11.281691, acc = 0.970000\n",
      "L = 11.276590, acc = 0.970000\n",
      "L = 11.271501, acc = 0.970000\n",
      "L = 11.266422, acc = 0.970000\n",
      "L = 11.261354, acc = 0.970000\n",
      "L = 11.256296, acc = 0.970000\n",
      "L = 11.251249, acc = 0.970000\n",
      "L = 11.246213, acc = 0.970000\n",
      "L = 11.241187, acc = 0.970000\n",
      "L = 11.236171, acc = 0.970000\n",
      "L = 11.231166, acc = 0.970000\n",
      "L = 11.226172, acc = 0.970000\n",
      "L = 11.221187, acc = 0.970000\n",
      "L = 11.216213, acc = 0.970000\n",
      "L = 11.211249, acc = 0.970000\n",
      "L = 11.206296, acc = 0.970000\n",
      "L = 11.201352, acc = 0.970000\n",
      "L = 11.196419, acc = 0.970000\n",
      "L = 11.191496, acc = 0.970000\n",
      "L = 11.186583, acc = 0.970000\n",
      "L = 11.181680, acc = 0.970000\n",
      "L = 11.176787, acc = 0.970000\n",
      "L = 11.171904, acc = 0.970000\n",
      "L = 11.167031, acc = 0.970000\n",
      "L = 11.162167, acc = 0.970000\n",
      "L = 11.157314, acc = 0.970000\n",
      "L = 11.152470, acc = 0.970000\n",
      "L = 11.147637, acc = 0.970000\n",
      "L = 11.142813, acc = 0.970000\n",
      "L = 11.137998, acc = 0.970000\n",
      "L = 11.133194, acc = 0.970000\n",
      "L = 11.128399, acc = 0.970000\n",
      "L = 11.123613, acc = 0.970000\n",
      "L = 11.118838, acc = 0.970000\n",
      "L = 11.114072, acc = 0.970000\n",
      "L = 11.109315, acc = 0.970000\n",
      "L = 11.104568, acc = 0.970000\n",
      "L = 11.099830, acc = 0.970000\n",
      "L = 11.095102, acc = 0.970000\n",
      "L = 11.090383, acc = 0.970000\n",
      "L = 11.085673, acc = 0.970000\n",
      "L = 11.080973, acc = 0.970000\n",
      "L = 11.076282, acc = 0.970000\n",
      "L = 11.071600, acc = 0.970000\n",
      "L = 11.066928, acc = 0.970000\n",
      "L = 11.062264, acc = 0.970000\n",
      "L = 11.057610, acc = 0.970000\n",
      "L = 11.052965, acc = 0.970000\n",
      "L = 11.048329, acc = 0.970000\n",
      "L = 11.043702, acc = 0.970000\n",
      "L = 11.039085, acc = 0.970000\n",
      "L = 11.034476, acc = 0.970000\n",
      "L = 11.029876, acc = 0.970000\n",
      "L = 11.025285, acc = 0.970000\n",
      "L = 11.020703, acc = 0.970000\n",
      "L = 11.016130, acc = 0.970000\n",
      "L = 11.011566, acc = 0.970000\n",
      "L = 11.007011, acc = 0.970000\n",
      "L = 11.002464, acc = 0.970000\n",
      "L = 10.997927, acc = 0.970000\n",
      "L = 10.993398, acc = 0.970000\n",
      "L = 10.988877, acc = 0.970000\n",
      "L = 10.984366, acc = 0.970000\n",
      "L = 10.979863, acc = 0.970000\n",
      "L = 10.975369, acc = 0.970000\n",
      "L = 10.970883, acc = 0.970000\n",
      "L = 10.966406, acc = 0.970000\n",
      "L = 10.961938, acc = 0.970000\n",
      "L = 10.957478, acc = 0.970000\n",
      "L = 10.953026, acc = 0.970000\n",
      "L = 10.948583, acc = 0.970000\n",
      "L = 10.944149, acc = 0.970000\n",
      "L = 10.939723, acc = 0.970000\n",
      "L = 10.935305, acc = 0.970000\n",
      "L = 10.930896, acc = 0.965000\n",
      "L = 10.926495, acc = 0.965000\n",
      "L = 10.922102, acc = 0.965000\n",
      "L = 10.917718, acc = 0.965000\n",
      "L = 10.913342, acc = 0.965000\n",
      "L = 10.908974, acc = 0.965000\n",
      "L = 10.904614, acc = 0.965000\n",
      "L = 10.900263, acc = 0.965000\n",
      "L = 10.895920, acc = 0.965000\n",
      "L = 10.891585, acc = 0.965000\n",
      "L = 10.887258, acc = 0.965000\n",
      "L = 10.882939, acc = 0.965000\n",
      "L = 10.878628, acc = 0.965000\n",
      "L = 10.874325, acc = 0.965000\n",
      "L = 10.870031, acc = 0.965000\n",
      "L = 10.865744, acc = 0.965000\n",
      "L = 10.861465, acc = 0.965000\n",
      "L = 10.857195, acc = 0.965000\n",
      "L = 10.852932, acc = 0.965000\n",
      "L = 10.848677, acc = 0.965000\n",
      "L = 10.844430, acc = 0.965000\n",
      "L = 10.840191, acc = 0.965000\n",
      "L = 10.835959, acc = 0.965000\n",
      "L = 10.831736, acc = 0.965000\n",
      "L = 10.827520, acc = 0.965000\n",
      "L = 10.823312, acc = 0.965000\n",
      "L = 10.819112, acc = 0.965000\n",
      "L = 10.814919, acc = 0.965000\n",
      "L = 10.810735, acc = 0.965000\n",
      "L = 10.806558, acc = 0.965000\n",
      "L = 10.802388, acc = 0.965000\n",
      "L = 10.798226, acc = 0.965000\n",
      "L = 10.794072, acc = 0.965000\n",
      "L = 10.789926, acc = 0.965000\n",
      "L = 10.785787, acc = 0.965000\n",
      "L = 10.781655, acc = 0.965000\n",
      "L = 10.777531, acc = 0.965000\n",
      "L = 10.773415, acc = 0.965000\n",
      "L = 10.769306, acc = 0.965000\n",
      "L = 10.765205, acc = 0.965000\n",
      "L = 10.761111, acc = 0.965000\n",
      "L = 10.757024, acc = 0.965000\n",
      "L = 10.752945, acc = 0.965000\n",
      "L = 10.748873, acc = 0.965000\n",
      "L = 10.744809, acc = 0.965000\n",
      "L = 10.740752, acc = 0.965000\n",
      "L = 10.736702, acc = 0.965000\n",
      "L = 10.732660, acc = 0.965000\n",
      "L = 10.728625, acc = 0.965000\n",
      "L = 10.724597, acc = 0.965000\n",
      "L = 10.720576, acc = 0.965000\n",
      "L = 10.716563, acc = 0.965000\n",
      "L = 10.712557, acc = 0.965000\n",
      "L = 10.708558, acc = 0.965000\n",
      "L = 10.704566, acc = 0.965000\n",
      "L = 10.700582, acc = 0.965000\n",
      "L = 10.696604, acc = 0.965000\n",
      "L = 10.692634, acc = 0.965000\n",
      "L = 10.688671, acc = 0.965000\n",
      "L = 10.684715, acc = 0.965000\n",
      "L = 10.680766, acc = 0.965000\n",
      "L = 10.676824, acc = 0.965000\n",
      "L = 10.672889, acc = 0.965000\n",
      "L = 10.668961, acc = 0.965000\n",
      "L = 10.665040, acc = 0.965000\n",
      "L = 10.661126, acc = 0.965000\n",
      "L = 10.657219, acc = 0.965000\n",
      "L = 10.653318, acc = 0.965000\n",
      "L = 10.649425, acc = 0.965000\n",
      "L = 10.645539, acc = 0.965000\n",
      "L = 10.641659, acc = 0.965000\n",
      "L = 10.637787, acc = 0.965000\n",
      "L = 10.633921, acc = 0.965000\n",
      "L = 10.630062, acc = 0.965000\n",
      "L = 10.626210, acc = 0.965000\n",
      "L = 10.622365, acc = 0.965000\n",
      "L = 10.618526, acc = 0.965000\n",
      "L = 10.614695, acc = 0.965000\n",
      "L = 10.610870, acc = 0.965000\n",
      "L = 10.607051, acc = 0.965000\n",
      "L = 10.603240, acc = 0.965000\n",
      "L = 10.599435, acc = 0.965000\n",
      "L = 10.595637, acc = 0.965000\n",
      "L = 10.591845, acc = 0.965000\n",
      "L = 10.588060, acc = 0.965000\n",
      "L = 10.584282, acc = 0.965000\n",
      "L = 10.580510, acc = 0.965000\n",
      "L = 10.576745, acc = 0.965000\n",
      "L = 10.572987, acc = 0.965000\n",
      "L = 10.569235, acc = 0.965000\n",
      "L = 10.565490, acc = 0.965000\n",
      "L = 10.561751, acc = 0.965000\n",
      "L = 10.558019, acc = 0.965000\n",
      "L = 10.554293, acc = 0.965000\n",
      "L = 10.550574, acc = 0.965000\n",
      "L = 10.546861, acc = 0.965000\n",
      "L = 10.543154, acc = 0.965000\n",
      "L = 10.539454, acc = 0.965000\n",
      "L = 10.535761, acc = 0.965000\n",
      "L = 10.532074, acc = 0.965000\n",
      "L = 10.528393, acc = 0.965000\n",
      "L = 10.524719, acc = 0.965000\n",
      "L = 10.521051, acc = 0.965000\n",
      "L = 10.517389, acc = 0.965000\n",
      "L = 10.513734, acc = 0.965000\n",
      "L = 10.510085, acc = 0.965000\n",
      "L = 10.506442, acc = 0.965000\n",
      "L = 10.502806, acc = 0.965000\n",
      "L = 10.499176, acc = 0.965000\n",
      "L = 10.495552, acc = 0.965000\n",
      "L = 10.491934, acc = 0.965000\n",
      "L = 10.488323, acc = 0.965000\n",
      "L = 10.484717, acc = 0.965000\n",
      "L = 10.481118, acc = 0.965000\n",
      "L = 10.477526, acc = 0.965000\n",
      "L = 10.473939, acc = 0.965000\n",
      "L = 10.470359, acc = 0.965000\n",
      "L = 10.466784, acc = 0.965000\n",
      "L = 10.463216, acc = 0.965000\n",
      "L = 10.459654, acc = 0.965000\n",
      "L = 10.456098, acc = 0.965000\n",
      "L = 10.452548, acc = 0.965000\n",
      "L = 10.449004, acc = 0.965000\n",
      "L = 10.445466, acc = 0.965000\n",
      "L = 10.441935, acc = 0.965000\n",
      "L = 10.438409, acc = 0.965000\n",
      "L = 10.434889, acc = 0.965000\n",
      "L = 10.431376, acc = 0.965000\n",
      "L = 10.427868, acc = 0.965000\n",
      "L = 10.424366, acc = 0.965000\n",
      "L = 10.420870, acc = 0.965000\n",
      "L = 10.417381, acc = 0.965000\n",
      "L = 10.413897, acc = 0.965000\n",
      "L = 10.410419, acc = 0.965000\n",
      "L = 10.406947, acc = 0.965000\n",
      "L = 10.403481, acc = 0.965000\n",
      "L = 10.400020, acc = 0.965000\n",
      "L = 10.396566, acc = 0.965000\n",
      "L = 10.393117, acc = 0.965000\n",
      "L = 10.389675, acc = 0.965000\n",
      "L = 10.386238, acc = 0.965000\n",
      "L = 10.382807, acc = 0.965000\n",
      "L = 10.379381, acc = 0.965000\n",
      "L = 10.375962, acc = 0.965000\n",
      "L = 10.372548, acc = 0.965000\n",
      "L = 10.369140, acc = 0.965000\n",
      "L = 10.365738, acc = 0.965000\n",
      "L = 10.362341, acc = 0.965000\n",
      "L = 10.358951, acc = 0.965000\n",
      "L = 10.355566, acc = 0.965000\n",
      "L = 10.352186, acc = 0.965000\n",
      "L = 10.348813, acc = 0.965000\n",
      "L = 10.345445, acc = 0.965000\n",
      "L = 10.342082, acc = 0.965000\n",
      "L = 10.338726, acc = 0.965000\n",
      "L = 10.335375, acc = 0.965000\n",
      "L = 10.332029, acc = 0.965000\n",
      "L = 10.328690, acc = 0.965000\n",
      "L = 10.325355, acc = 0.965000\n",
      "L = 10.322027, acc = 0.965000\n",
      "L = 10.318704, acc = 0.965000\n",
      "L = 10.315386, acc = 0.965000\n",
      "L = 10.312074, acc = 0.965000\n",
      "L = 10.308768, acc = 0.965000\n",
      "L = 10.305467, acc = 0.965000\n",
      "L = 10.302172, acc = 0.965000\n",
      "L = 10.298882, acc = 0.965000\n",
      "L = 10.295597, acc = 0.965000\n",
      "L = 10.292319, acc = 0.965000\n",
      "L = 10.289045, acc = 0.965000\n",
      "L = 10.285777, acc = 0.965000\n",
      "L = 10.282515, acc = 0.965000\n",
      "L = 10.279258, acc = 0.965000\n",
      "L = 10.276006, acc = 0.965000\n",
      "L = 10.272760, acc = 0.965000\n",
      "L = 10.269519, acc = 0.965000\n",
      "L = 10.266283, acc = 0.965000\n",
      "L = 10.263053, acc = 0.965000\n",
      "L = 10.259828, acc = 0.965000\n",
      "L = 10.256609, acc = 0.965000\n",
      "L = 10.253395, acc = 0.965000\n",
      "L = 10.250186, acc = 0.965000\n",
      "L = 10.246983, acc = 0.965000\n",
      "L = 10.243785, acc = 0.965000\n",
      "L = 10.240592, acc = 0.965000\n",
      "L = 10.237404, acc = 0.965000\n",
      "L = 10.234222, acc = 0.965000\n",
      "L = 10.231045, acc = 0.965000\n",
      "L = 10.227873, acc = 0.965000\n",
      "L = 10.224707, acc = 0.965000\n",
      "L = 10.221545, acc = 0.965000\n",
      "L = 10.218389, acc = 0.965000\n",
      "L = 10.215238, acc = 0.965000\n",
      "L = 10.212092, acc = 0.965000\n",
      "L = 10.208952, acc = 0.965000\n",
      "L = 10.205816, acc = 0.965000\n",
      "L = 10.202686, acc = 0.965000\n",
      "L = 10.199561, acc = 0.965000\n",
      "L = 10.196441, acc = 0.965000\n",
      "L = 10.193326, acc = 0.965000\n",
      "L = 10.190216, acc = 0.965000\n",
      "L = 10.187112, acc = 0.965000\n",
      "L = 10.184012, acc = 0.965000\n",
      "L = 10.180918, acc = 0.965000\n",
      "L = 10.177828, acc = 0.965000\n",
      "L = 10.174744, acc = 0.965000\n",
      "L = 10.171665, acc = 0.965000\n",
      "L = 10.168591, acc = 0.965000\n",
      "L = 10.165521, acc = 0.965000\n",
      "L = 10.162457, acc = 0.965000\n",
      "L = 10.159398, acc = 0.965000\n",
      "L = 10.156344, acc = 0.965000\n",
      "L = 10.153294, acc = 0.965000\n",
      "L = 10.150250, acc = 0.965000\n",
      "L = 10.147211, acc = 0.965000\n",
      "L = 10.144176, acc = 0.965000\n",
      "L = 10.141147, acc = 0.965000\n",
      "L = 10.138122, acc = 0.965000\n",
      "L = 10.135103, acc = 0.965000\n",
      "L = 10.132088, acc = 0.965000\n",
      "L = 10.129078, acc = 0.965000\n",
      "L = 10.126073, acc = 0.965000\n",
      "L = 10.123073, acc = 0.965000\n",
      "L = 10.120078, acc = 0.965000\n",
      "L = 10.117088, acc = 0.965000\n",
      "L = 10.114103, acc = 0.965000\n",
      "L = 10.111122, acc = 0.965000\n",
      "L = 10.108146, acc = 0.965000\n",
      "L = 10.105175, acc = 0.965000\n",
      "L = 10.102209, acc = 0.965000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 10.099248, acc = 0.965000\n",
      "L = 10.096291, acc = 0.965000\n",
      "L = 10.093339, acc = 0.965000\n",
      "L = 10.090392, acc = 0.965000\n",
      "L = 10.087450, acc = 0.965000\n",
      "L = 10.084512, acc = 0.965000\n",
      "L = 10.081579, acc = 0.965000\n",
      "L = 10.078651, acc = 0.965000\n",
      "L = 10.075728, acc = 0.965000\n",
      "L = 10.072809, acc = 0.965000\n",
      "L = 10.069895, acc = 0.965000\n",
      "L = 10.066986, acc = 0.965000\n",
      "L = 10.064081, acc = 0.965000\n",
      "L = 10.061181, acc = 0.965000\n",
      "L = 10.058286, acc = 0.965000\n",
      "L = 10.055395, acc = 0.965000\n",
      "L = 10.052509, acc = 0.965000\n",
      "L = 10.049628, acc = 0.965000\n",
      "L = 10.046751, acc = 0.965000\n",
      "L = 10.043879, acc = 0.965000\n",
      "L = 10.041011, acc = 0.965000\n",
      "L = 10.038148, acc = 0.965000\n",
      "L = 10.035290, acc = 0.965000\n",
      "L = 10.032436, acc = 0.965000\n",
      "L = 10.029586, acc = 0.965000\n",
      "L = 10.026742, acc = 0.965000\n",
      "L = 10.023901, acc = 0.965000\n",
      "L = 10.021066, acc = 0.965000\n",
      "L = 10.018234, acc = 0.965000\n",
      "L = 10.015408, acc = 0.965000\n",
      "L = 10.012586, acc = 0.965000\n",
      "L = 10.009768, acc = 0.965000\n",
      "L = 10.006955, acc = 0.965000\n",
      "L = 10.004146, acc = 0.965000\n",
      "L = 10.001342, acc = 0.965000\n",
      "L = 9.998542, acc = 0.965000\n",
      "L = 9.995746, acc = 0.965000\n",
      "L = 9.992955, acc = 0.965000\n",
      "L = 9.990169, acc = 0.965000\n",
      "L = 9.987387, acc = 0.965000\n",
      "L = 9.984609, acc = 0.965000\n",
      "L = 9.981835, acc = 0.965000\n",
      "L = 9.979066, acc = 0.965000\n",
      "L = 9.976302, acc = 0.965000\n",
      "L = 9.973542, acc = 0.965000\n",
      "L = 9.970786, acc = 0.965000\n",
      "L = 9.968034, acc = 0.965000\n",
      "L = 9.965287, acc = 0.965000\n",
      "L = 9.962544, acc = 0.965000\n",
      "L = 9.959806, acc = 0.965000\n",
      "L = 9.957071, acc = 0.965000\n",
      "L = 9.954342, acc = 0.965000\n",
      "L = 9.951616, acc = 0.965000\n",
      "L = 9.948895, acc = 0.965000\n",
      "L = 9.946177, acc = 0.965000\n",
      "L = 9.943465, acc = 0.970000\n",
      "L = 9.940756, acc = 0.970000\n",
      "L = 9.938052, acc = 0.970000\n",
      "L = 9.935352, acc = 0.970000\n",
      "L = 9.932656, acc = 0.970000\n",
      "L = 9.929964, acc = 0.970000\n",
      "L = 9.927277, acc = 0.970000\n",
      "L = 9.924594, acc = 0.970000\n",
      "L = 9.921915, acc = 0.970000\n",
      "L = 9.919240, acc = 0.970000\n",
      "L = 9.916569, acc = 0.970000\n",
      "L = 9.913903, acc = 0.970000\n",
      "L = 9.911240, acc = 0.970000\n",
      "L = 9.908582, acc = 0.970000\n",
      "L = 9.905928, acc = 0.970000\n",
      "L = 9.903278, acc = 0.970000\n",
      "L = 9.900632, acc = 0.970000\n",
      "L = 9.897991, acc = 0.970000\n",
      "L = 9.895353, acc = 0.970000\n",
      "L = 9.892720, acc = 0.970000\n",
      "L = 9.890090, acc = 0.970000\n",
      "L = 9.887465, acc = 0.970000\n",
      "L = 9.884844, acc = 0.970000\n",
      "L = 9.882227, acc = 0.970000\n",
      "L = 9.879614, acc = 0.970000\n",
      "L = 9.877004, acc = 0.970000\n",
      "L = 9.874399, acc = 0.970000\n",
      "L = 9.871798, acc = 0.970000\n",
      "L = 9.869201, acc = 0.970000\n",
      "L = 9.866608, acc = 0.970000\n",
      "L = 9.864019, acc = 0.970000\n",
      "L = 9.861434, acc = 0.970000\n",
      "L = 9.858853, acc = 0.970000\n",
      "L = 9.856276, acc = 0.970000\n",
      "L = 9.853703, acc = 0.970000\n",
      "L = 9.851134, acc = 0.970000\n",
      "L = 9.848569, acc = 0.970000\n",
      "L = 9.846008, acc = 0.970000\n",
      "L = 9.843450, acc = 0.970000\n",
      "L = 9.840897, acc = 0.970000\n",
      "L = 9.838348, acc = 0.970000\n",
      "L = 9.835802, acc = 0.970000\n",
      "L = 9.833261, acc = 0.970000\n",
      "L = 9.830723, acc = 0.970000\n",
      "L = 9.828189, acc = 0.970000\n",
      "L = 9.825659, acc = 0.970000\n",
      "L = 9.823133, acc = 0.970000\n",
      "L = 9.820611, acc = 0.970000\n",
      "L = 9.818092, acc = 0.970000\n",
      "L = 9.815578, acc = 0.970000\n",
      "L = 9.813067, acc = 0.970000\n",
      "L = 9.810560, acc = 0.970000\n",
      "L = 9.808057, acc = 0.970000\n",
      "L = 9.805558, acc = 0.970000\n",
      "L = 9.803062, acc = 0.970000\n",
      "L = 9.800570, acc = 0.970000\n",
      "L = 9.798082, acc = 0.970000\n",
      "L = 9.795598, acc = 0.970000\n",
      "L = 9.793118, acc = 0.970000\n",
      "L = 9.790641, acc = 0.970000\n",
      "L = 9.788168, acc = 0.970000\n",
      "L = 9.785699, acc = 0.970000\n",
      "L = 9.783234, acc = 0.970000\n",
      "L = 9.780772, acc = 0.970000\n",
      "L = 9.778314, acc = 0.970000\n",
      "L = 9.775860, acc = 0.970000\n",
      "L = 9.773410, acc = 0.970000\n",
      "L = 9.770963, acc = 0.970000\n",
      "L = 9.768520, acc = 0.970000\n",
      "L = 9.766080, acc = 0.970000\n",
      "L = 9.763645, acc = 0.970000\n",
      "L = 9.761213, acc = 0.970000\n",
      "L = 9.758784, acc = 0.970000\n",
      "L = 9.756359, acc = 0.970000\n",
      "L = 9.753938, acc = 0.970000\n",
      "L = 9.751521, acc = 0.970000\n",
      "L = 9.749107, acc = 0.970000\n",
      "L = 9.746696, acc = 0.970000\n",
      "L = 9.744290, acc = 0.970000\n",
      "L = 9.741887, acc = 0.970000\n",
      "L = 9.739487, acc = 0.970000\n",
      "L = 9.737091, acc = 0.970000\n",
      "L = 9.734699, acc = 0.970000\n",
      "L = 9.732310, acc = 0.970000\n",
      "L = 9.729925, acc = 0.970000\n",
      "L = 9.727544, acc = 0.970000\n",
      "L = 9.725166, acc = 0.970000\n",
      "L = 9.722791, acc = 0.970000\n",
      "L = 9.720420, acc = 0.970000\n",
      "L = 9.718053, acc = 0.970000\n",
      "L = 9.715689, acc = 0.970000\n",
      "L = 9.713328, acc = 0.970000\n",
      "L = 9.710971, acc = 0.970000\n",
      "L = 9.708618, acc = 0.970000\n",
      "L = 9.706268, acc = 0.970000\n",
      "L = 9.703921, acc = 0.970000\n",
      "L = 9.701578, acc = 0.970000\n",
      "L = 9.699239, acc = 0.970000\n",
      "L = 9.696903, acc = 0.970000\n",
      "L = 9.694570, acc = 0.970000\n",
      "L = 9.692241, acc = 0.970000\n",
      "L = 9.689915, acc = 0.970000\n",
      "L = 9.687593, acc = 0.970000\n",
      "L = 9.685274, acc = 0.970000\n",
      "L = 9.682959, acc = 0.970000\n",
      "L = 9.680647, acc = 0.970000\n",
      "L = 9.678338, acc = 0.970000\n",
      "L = 9.676033, acc = 0.970000\n",
      "L = 9.673731, acc = 0.970000\n",
      "L = 9.671432, acc = 0.970000\n",
      "L = 9.669137, acc = 0.970000\n",
      "L = 9.666845, acc = 0.970000\n",
      "L = 9.664557, acc = 0.970000\n",
      "L = 9.662272, acc = 0.970000\n",
      "L = 9.659990, acc = 0.970000\n",
      "L = 9.657712, acc = 0.970000\n",
      "L = 9.655437, acc = 0.970000\n",
      "L = 9.653165, acc = 0.970000\n",
      "L = 9.650897, acc = 0.970000\n",
      "L = 9.648631, acc = 0.970000\n",
      "L = 9.646370, acc = 0.970000\n",
      "L = 9.644111, acc = 0.970000\n",
      "L = 9.641856, acc = 0.970000\n",
      "L = 9.639604, acc = 0.970000\n",
      "L = 9.637355, acc = 0.970000\n",
      "L = 9.635110, acc = 0.970000\n",
      "L = 9.632868, acc = 0.970000\n",
      "L = 9.630629, acc = 0.970000\n",
      "L = 9.628393, acc = 0.970000\n",
      "L = 9.626160, acc = 0.970000\n",
      "L = 9.623931, acc = 0.970000\n",
      "L = 9.621705, acc = 0.970000\n",
      "L = 9.619482, acc = 0.970000\n",
      "L = 9.617263, acc = 0.970000\n",
      "L = 9.615046, acc = 0.970000\n",
      "L = 9.612833, acc = 0.970000\n",
      "L = 9.610623, acc = 0.970000\n",
      "L = 9.608416, acc = 0.970000\n",
      "L = 9.606213, acc = 0.970000\n",
      "L = 9.604012, acc = 0.970000\n",
      "L = 9.601815, acc = 0.970000\n",
      "L = 9.599621, acc = 0.970000\n",
      "L = 9.597430, acc = 0.970000\n",
      "L = 9.595242, acc = 0.970000\n",
      "L = 9.593057, acc = 0.970000\n",
      "L = 9.590876, acc = 0.970000\n",
      "L = 9.588697, acc = 0.970000\n",
      "L = 9.586522, acc = 0.970000\n",
      "L = 9.584349, acc = 0.970000\n",
      "L = 9.582180, acc = 0.970000\n",
      "L = 9.580014, acc = 0.970000\n",
      "L = 9.577851, acc = 0.970000\n",
      "L = 9.575691, acc = 0.970000\n",
      "L = 9.573534, acc = 0.970000\n",
      "L = 9.571381, acc = 0.970000\n",
      "L = 9.569230, acc = 0.970000\n",
      "L = 9.567082, acc = 0.970000\n",
      "L = 9.564937, acc = 0.970000\n",
      "L = 9.562796, acc = 0.970000\n",
      "L = 9.560657, acc = 0.970000\n",
      "L = 9.558522, acc = 0.970000\n",
      "L = 9.556389, acc = 0.970000\n",
      "L = 9.554260, acc = 0.970000\n",
      "L = 9.552133, acc = 0.970000\n",
      "L = 9.550010, acc = 0.970000\n",
      "L = 9.547889, acc = 0.970000\n",
      "L = 9.545772, acc = 0.970000\n",
      "L = 9.543657, acc = 0.970000\n",
      "L = 9.541546, acc = 0.970000\n",
      "L = 9.539437, acc = 0.970000\n",
      "L = 9.537332, acc = 0.970000\n",
      "L = 9.535229, acc = 0.970000\n",
      "L = 9.533129, acc = 0.970000\n",
      "L = 9.531032, acc = 0.970000\n",
      "L = 9.528939, acc = 0.970000\n",
      "L = 9.526848, acc = 0.970000\n",
      "L = 9.524760, acc = 0.970000\n",
      "L = 9.522675, acc = 0.970000\n",
      "L = 9.520592, acc = 0.970000\n",
      "L = 9.518513, acc = 0.970000\n",
      "L = 9.516437, acc = 0.970000\n",
      "L = 9.514363, acc = 0.970000\n",
      "L = 9.512293, acc = 0.970000\n",
      "L = 9.510225, acc = 0.970000\n",
      "L = 9.508160, acc = 0.970000\n",
      "L = 9.506098, acc = 0.970000\n",
      "L = 9.504039, acc = 0.970000\n",
      "L = 9.501983, acc = 0.970000\n",
      "L = 9.499930, acc = 0.970000\n",
      "L = 9.497879, acc = 0.970000\n",
      "L = 9.495832, acc = 0.970000\n",
      "L = 9.493787, acc = 0.975000\n",
      "L = 9.491745, acc = 0.975000\n",
      "L = 9.489706, acc = 0.975000\n",
      "L = 9.487669, acc = 0.975000\n",
      "L = 9.485636, acc = 0.975000\n",
      "L = 9.483605, acc = 0.975000\n",
      "L = 9.481577, acc = 0.975000\n",
      "L = 9.479552, acc = 0.975000\n",
      "L = 9.477529, acc = 0.975000\n",
      "L = 9.475510, acc = 0.975000\n",
      "L = 9.473493, acc = 0.975000\n",
      "L = 9.471479, acc = 0.975000\n",
      "L = 9.469467, acc = 0.975000\n",
      "L = 9.467459, acc = 0.975000\n",
      "L = 9.465453, acc = 0.975000\n",
      "L = 9.463450, acc = 0.975000\n",
      "L = 9.461450, acc = 0.975000\n",
      "L = 9.459452, acc = 0.975000\n",
      "L = 9.457457, acc = 0.975000\n",
      "L = 9.455465, acc = 0.975000\n",
      "L = 9.453475, acc = 0.975000\n",
      "L = 9.451489, acc = 0.975000\n",
      "L = 9.449505, acc = 0.975000\n",
      "L = 9.447523, acc = 0.975000\n",
      "L = 9.445545, acc = 0.975000\n",
      "L = 9.443569, acc = 0.975000\n",
      "L = 9.441595, acc = 0.975000\n",
      "L = 9.439625, acc = 0.975000\n",
      "L = 9.437657, acc = 0.975000\n",
      "L = 9.435692, acc = 0.975000\n",
      "L = 9.433729, acc = 0.975000\n"
     ]
    }
   ],
   "source": [
    "# use the NN model and training\n",
    "nn = NN_Model([2, 6, 4, 2])\n",
    "nn.init_weight()\n",
    "nn.backpropagation(X, t, 2000)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# predict results & plot results\n",
    "y_res  = nn.forward(X)\n",
    "y_pred = np.argmax(y_res, axis=1)\n",
    "\n",
    "# plot data\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y, cmap=plt.cm.Spectral)\n",
    "plt.title(\"ground truth\")\n",
    "plt.show()\n",
    "\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y_pred, cmap=plt.cm.Spectral)\n",
    "plt.title(\"predicted\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 10. 深入分析与问题"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.01102277 0.98892257]\n",
      " [0.13689246 0.86620671]\n",
      " [0.97904664 0.02132821]\n",
      " [0.01163523 0.98829983]\n",
      " [0.00717948 0.99279357]\n",
      " [0.95281465 0.04607736]\n",
      " [0.01748735 0.98260651]\n",
      " [0.97215654 0.0271742 ]\n",
      " [0.03769688 0.96206663]]\n"
     ]
    }
   ],
   "source": [
    "# print some results\n",
    "\n",
    "print(y_res[1:10, :])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**问题**\n",
    "1. 我们希望得到的每个类别的概率，如何实现？\n",
    "2. 如何做多分类问题？\n",
    "3. 如何能让神经网络更快的训练好？\n",
    "4. 如何更好的构建网络的类定义和接口设计，从而让神经网络的类支持更多的类型的处理层？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## References\n",
    "\n",
    "* [零基础入门深度学习(3) - 神经网络和反向传播算法](https://www.zybuluo.com/hanbingtao/note/476663)\n",
    "* [Neural Network Using Python and Numpy](https://www.python-course.eu/neural_networks_with_python_numpy.php)\n",
    "* http://www.cedar.buffalo.edu/%7Esrihari/CSE574/Chap5/Chap5.3-BackProp.pdf\n",
    "* https://mattmazur.com/2015/03/17/a-step-by-step-backpropagation-example/\n"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}