machinelearning_notebook/nn/mlp_bp.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 多层神经网络和反向传播\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 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": [
    "## 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。最左边的层叫做输入层，负责接收输入数据；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。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": [
    "## 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",
    "\n",
    "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",
    "如上图，输入层有三个节点，我们将其依次编号为1、2、3；隐藏层的4个节点，编号依次为4、5、6、7；最后输出层的两个节点编号为8、9。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",
    "\n",
    "\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，你偏非要把$x_1$赋值给节点2也是完全没有问题的，但这样除了把自己弄晕之外，并没有什么价值。\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",
    "\n",
    "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",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 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（当然逼格也更高），我们先来看看隐藏层的矩阵表示。\n",
    "\n",
    "首先我们把隐藏层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",
    "\n",
    "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"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 神经网络的训练 - 反向传播算法\n",
    "\n",
    "现在，我们需要知道一个神经网络的每个连接上的权值是如何得到的。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",
    "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。\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：\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": [
    "### 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": [
    "### 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": [
    "###  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": [
    "## Program"
   ]
  },
  {
   "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": 3,
   "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 = 4                # 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",
    "# defin sigmod & its derivate function\n",
    "def sigmod(X):\n",
    "    return 1.0/(1+np.exp(-X))\n",
    "\n",
    "def sigmod_derivative(X):\n",
    "    f = sigmod(X)\n",
    "    return f*(1-f)\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": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch [   0] L = 85.451119, acc = 0.815000\n",
      "epoch [   1] L = 83.419661, acc = 0.805000\n",
      "epoch [   2] L = 81.672674, acc = 0.785000\n",
      "epoch [   3] L = 80.109531, acc = 0.790000\n",
      "epoch [   4] L = 78.664310, acc = 0.800000\n",
      "epoch [   5] L = 77.297470, acc = 0.795000\n",
      "epoch [   6] L = 75.986789, acc = 0.780000\n",
      "epoch [   7] L = 74.720510, acc = 0.790000\n",
      "epoch [   8] L = 73.492869, acc = 0.795000\n",
      "epoch [   9] L = 72.301379, acc = 0.795000\n",
      "epoch [  10] L = 71.145261, acc = 0.795000\n",
      "epoch [  11] L = 70.024545, acc = 0.795000\n",
      "epoch [  12] L = 68.939583, acc = 0.795000\n",
      "epoch [  13] L = 67.890772, acc = 0.795000\n",
      "epoch [  14] L = 66.878412, acc = 0.795000\n",
      "epoch [  15] L = 65.902642, acc = 0.795000\n",
      "epoch [  16] L = 64.963401, acc = 0.800000\n",
      "epoch [  17] L = 64.060427, acc = 0.800000\n",
      "epoch [  18] L = 63.193267, acc = 0.800000\n",
      "epoch [  19] L = 62.361290, acc = 0.800000\n",
      "epoch [  20] L = 61.563710, acc = 0.800000\n",
      "epoch [  21] L = 60.799612, acc = 0.800000\n",
      "epoch [  22] L = 60.067969, acc = 0.800000\n",
      "epoch [  23] L = 59.367673, acc = 0.800000\n",
      "epoch [  24] L = 58.697554, acc = 0.800000\n",
      "epoch [  25] L = 58.056399, acc = 0.800000\n",
      "epoch [  26] L = 57.442972, acc = 0.805000\n",
      "epoch [  27] L = 56.856029, acc = 0.810000\n",
      "epoch [  28] L = 56.294331, acc = 0.815000\n",
      "epoch [  29] L = 55.756659, acc = 0.815000\n",
      "epoch [  30] L = 55.241816, acc = 0.820000\n",
      "epoch [  31] L = 54.748641, acc = 0.820000\n",
      "epoch [  32] L = 54.276013, acc = 0.820000\n",
      "epoch [  33] L = 53.822853, acc = 0.820000\n",
      "epoch [  34] L = 53.388127, acc = 0.820000\n",
      "epoch [  35] L = 52.970853, acc = 0.825000\n",
      "epoch [  36] L = 52.570093, acc = 0.825000\n",
      "epoch [  37] L = 52.184962, acc = 0.825000\n",
      "epoch [  38] L = 51.814620, acc = 0.825000\n",
      "epoch [  39] L = 51.458277, acc = 0.825000\n",
      "epoch [  40] L = 51.115188, acc = 0.825000\n",
      "epoch [  41] L = 50.784653, acc = 0.825000\n",
      "epoch [  42] L = 50.466013, acc = 0.825000\n",
      "epoch [  43] L = 50.158651, acc = 0.825000\n",
      "epoch [  44] L = 49.861989, acc = 0.825000\n",
      "epoch [  45] L = 49.575485, acc = 0.825000\n",
      "epoch [  46] L = 49.298632, acc = 0.825000\n",
      "epoch [  47] L = 49.030954, acc = 0.825000\n",
      "epoch [  48] L = 48.772008, acc = 0.825000\n",
      "epoch [  49] L = 48.521378, acc = 0.825000\n",
      "epoch [  50] L = 48.278674, acc = 0.825000\n",
      "epoch [  51] L = 48.043534, acc = 0.825000\n",
      "epoch [  52] L = 47.815615, acc = 0.825000\n",
      "epoch [  53] L = 47.594598, acc = 0.825000\n",
      "epoch [  54] L = 47.380186, acc = 0.825000\n",
      "epoch [  55] L = 47.172098, acc = 0.825000\n",
      "epoch [  56] L = 46.970070, acc = 0.825000\n",
      "epoch [  57] L = 46.773857, acc = 0.825000\n",
      "epoch [  58] L = 46.583226, acc = 0.825000\n",
      "epoch [  59] L = 46.397960, acc = 0.825000\n",
      "epoch [  60] L = 46.217854, acc = 0.825000\n",
      "epoch [  61] L = 46.042714, acc = 0.825000\n",
      "epoch [  62] L = 45.872360, acc = 0.825000\n",
      "epoch [  63] L = 45.706618, acc = 0.825000\n",
      "epoch [  64] L = 45.545327, acc = 0.825000\n",
      "epoch [  65] L = 45.388334, acc = 0.825000\n",
      "epoch [  66] L = 45.235492, acc = 0.825000\n",
      "epoch [  67] L = 45.086664, acc = 0.830000\n",
      "epoch [  68] L = 44.941719, acc = 0.830000\n",
      "epoch [  69] L = 44.800533, acc = 0.830000\n",
      "epoch [  70] L = 44.662986, acc = 0.830000\n",
      "epoch [  71] L = 44.528966, acc = 0.830000\n",
      "epoch [  72] L = 44.398366, acc = 0.830000\n",
      "epoch [  73] L = 44.271082, acc = 0.830000\n",
      "epoch [  74] L = 44.147015, acc = 0.835000\n",
      "epoch [  75] L = 44.026072, acc = 0.840000\n",
      "epoch [  76] L = 43.908161, acc = 0.840000\n",
      "epoch [  77] L = 43.793197, acc = 0.840000\n",
      "epoch [  78] L = 43.681095, acc = 0.835000\n",
      "epoch [  79] L = 43.571774, acc = 0.835000\n",
      "epoch [  80] L = 43.465159, acc = 0.835000\n",
      "epoch [  81] L = 43.361173, acc = 0.835000\n",
      "epoch [  82] L = 43.259745, acc = 0.835000\n",
      "epoch [  83] L = 43.160805, acc = 0.835000\n",
      "epoch [  84] L = 43.064286, acc = 0.835000\n",
      "epoch [  85] L = 42.970123, acc = 0.835000\n",
      "epoch [  86] L = 42.878253, acc = 0.835000\n",
      "epoch [  87] L = 42.788615, acc = 0.835000\n",
      "epoch [  88] L = 42.701150, acc = 0.835000\n",
      "epoch [  89] L = 42.615800, acc = 0.835000\n",
      "epoch [  90] L = 42.532509, acc = 0.835000\n",
      "epoch [  91] L = 42.451224, acc = 0.835000\n",
      "epoch [  92] L = 42.371892, acc = 0.835000\n",
      "epoch [  93] L = 42.294462, acc = 0.835000\n",
      "epoch [  94] L = 42.218884, acc = 0.835000\n",
      "epoch [  95] L = 42.145111, acc = 0.835000\n",
      "epoch [  96] L = 42.073095, acc = 0.835000\n",
      "epoch [  97] L = 42.002790, acc = 0.835000\n",
      "epoch [  98] L = 41.934152, acc = 0.835000\n",
      "epoch [  99] L = 41.867138, acc = 0.830000\n",
      "epoch [ 100] L = 41.801707, acc = 0.830000\n",
      "epoch [ 101] L = 41.737816, acc = 0.830000\n",
      "epoch [ 102] L = 41.675426, acc = 0.830000\n",
      "epoch [ 103] L = 41.614499, acc = 0.830000\n",
      "epoch [ 104] L = 41.554997, acc = 0.830000\n",
      "epoch [ 105] L = 41.496882, acc = 0.830000\n",
      "epoch [ 106] L = 41.440119, acc = 0.835000\n",
      "epoch [ 107] L = 41.384673, acc = 0.835000\n",
      "epoch [ 108] L = 41.330511, acc = 0.835000\n",
      "epoch [ 109] L = 41.277599, acc = 0.835000\n",
      "epoch [ 110] L = 41.225904, acc = 0.835000\n",
      "epoch [ 111] L = 41.175397, acc = 0.835000\n",
      "epoch [ 112] L = 41.126046, acc = 0.835000\n",
      "epoch [ 113] L = 41.077821, acc = 0.835000\n",
      "epoch [ 114] L = 41.030694, acc = 0.835000\n",
      "epoch [ 115] L = 40.984636, acc = 0.835000\n",
      "epoch [ 116] L = 40.939620, acc = 0.835000\n",
      "epoch [ 117] L = 40.895620, acc = 0.835000\n",
      "epoch [ 118] L = 40.852609, acc = 0.835000\n",
      "epoch [ 119] L = 40.810563, acc = 0.835000\n",
      "epoch [ 120] L = 40.769456, acc = 0.835000\n",
      "epoch [ 121] L = 40.729265, acc = 0.835000\n",
      "epoch [ 122] L = 40.689966, acc = 0.835000\n",
      "epoch [ 123] L = 40.651537, acc = 0.835000\n",
      "epoch [ 124] L = 40.613956, acc = 0.835000\n",
      "epoch [ 125] L = 40.577201, acc = 0.835000\n",
      "epoch [ 126] L = 40.541251, acc = 0.835000\n",
      "epoch [ 127] L = 40.506087, acc = 0.835000\n",
      "epoch [ 128] L = 40.471688, acc = 0.840000\n",
      "epoch [ 129] L = 40.438035, acc = 0.840000\n",
      "epoch [ 130] L = 40.405110, acc = 0.840000\n",
      "epoch [ 131] L = 40.372893, acc = 0.840000\n",
      "epoch [ 132] L = 40.341369, acc = 0.840000\n",
      "epoch [ 133] L = 40.310518, acc = 0.840000\n",
      "epoch [ 134] L = 40.280326, acc = 0.840000\n",
      "epoch [ 135] L = 40.250774, acc = 0.840000\n",
      "epoch [ 136] L = 40.221848, acc = 0.840000\n",
      "epoch [ 137] L = 40.193532, acc = 0.840000\n",
      "epoch [ 138] L = 40.165810, acc = 0.840000\n",
      "epoch [ 139] L = 40.138669, acc = 0.840000\n",
      "epoch [ 140] L = 40.112095, acc = 0.840000\n",
      "epoch [ 141] L = 40.086072, acc = 0.840000\n",
      "epoch [ 142] L = 40.060589, acc = 0.840000\n",
      "epoch [ 143] L = 40.035631, acc = 0.840000\n",
      "epoch [ 144] L = 40.011186, acc = 0.840000\n",
      "epoch [ 145] L = 39.987242, acc = 0.840000\n",
      "epoch [ 146] L = 39.963787, acc = 0.840000\n",
      "epoch [ 147] L = 39.940808, acc = 0.845000\n",
      "epoch [ 148] L = 39.918295, acc = 0.845000\n",
      "epoch [ 149] L = 39.896236, acc = 0.850000\n",
      "epoch [ 150] L = 39.874621, acc = 0.850000\n",
      "epoch [ 151] L = 39.853439, acc = 0.850000\n",
      "epoch [ 152] L = 39.832680, acc = 0.850000\n",
      "epoch [ 153] L = 39.812334, acc = 0.850000\n",
      "epoch [ 154] L = 39.792390, acc = 0.850000\n",
      "epoch [ 155] L = 39.772841, acc = 0.850000\n",
      "epoch [ 156] L = 39.753676, acc = 0.850000\n",
      "epoch [ 157] L = 39.734886, acc = 0.850000\n",
      "epoch [ 158] L = 39.716464, acc = 0.850000\n",
      "epoch [ 159] L = 39.698399, acc = 0.855000\n",
      "epoch [ 160] L = 39.680685, acc = 0.855000\n",
      "epoch [ 161] L = 39.663313, acc = 0.855000\n",
      "epoch [ 162] L = 39.646275, acc = 0.855000\n",
      "epoch [ 163] L = 39.629564, acc = 0.855000\n",
      "epoch [ 164] L = 39.613172, acc = 0.855000\n",
      "epoch [ 165] L = 39.597092, acc = 0.855000\n",
      "epoch [ 166] L = 39.581317, acc = 0.855000\n",
      "epoch [ 167] L = 39.565840, acc = 0.855000\n",
      "epoch [ 168] L = 39.550654, acc = 0.855000\n",
      "epoch [ 169] L = 39.535753, acc = 0.855000\n",
      "epoch [ 170] L = 39.521131, acc = 0.855000\n",
      "epoch [ 171] L = 39.506780, acc = 0.855000\n",
      "epoch [ 172] L = 39.492697, acc = 0.855000\n",
      "epoch [ 173] L = 39.478873, acc = 0.855000\n",
      "epoch [ 174] L = 39.465304, acc = 0.855000\n",
      "epoch [ 175] L = 39.451985, acc = 0.855000\n",
      "epoch [ 176] L = 39.438909, acc = 0.855000\n",
      "epoch [ 177] L = 39.426071, acc = 0.855000\n",
      "epoch [ 178] L = 39.413467, acc = 0.855000\n",
      "epoch [ 179] L = 39.401090, acc = 0.855000\n",
      "epoch [ 180] L = 39.388937, acc = 0.855000\n",
      "epoch [ 181] L = 39.377003, acc = 0.855000\n",
      "epoch [ 182] L = 39.365282, acc = 0.855000\n",
      "epoch [ 183] L = 39.353770, acc = 0.855000\n",
      "epoch [ 184] L = 39.342463, acc = 0.855000\n",
      "epoch [ 185] L = 39.331357, acc = 0.855000\n",
      "epoch [ 186] L = 39.320446, acc = 0.855000\n",
      "epoch [ 187] L = 39.309728, acc = 0.855000\n",
      "epoch [ 188] L = 39.299198, acc = 0.855000\n",
      "epoch [ 189] L = 39.288852, acc = 0.855000\n",
      "epoch [ 190] L = 39.278686, acc = 0.855000\n",
      "epoch [ 191] L = 39.268696, acc = 0.855000\n",
      "epoch [ 192] L = 39.258880, acc = 0.855000\n",
      "epoch [ 193] L = 39.249233, acc = 0.855000\n",
      "epoch [ 194] L = 39.239751, acc = 0.855000\n",
      "epoch [ 195] L = 39.230432, acc = 0.855000\n",
      "epoch [ 196] L = 39.221272, acc = 0.855000\n",
      "epoch [ 197] L = 39.212268, acc = 0.855000\n",
      "epoch [ 198] L = 39.203417, acc = 0.855000\n",
      "epoch [ 199] L = 39.194715, acc = 0.855000\n",
      "epoch [ 200] L = 39.186160, acc = 0.855000\n",
      "epoch [ 201] L = 39.177749, acc = 0.855000\n",
      "epoch [ 202] L = 39.169478, acc = 0.855000\n",
      "epoch [ 203] L = 39.161345, acc = 0.855000\n",
      "epoch [ 204] L = 39.153348, acc = 0.855000\n",
      "epoch [ 205] L = 39.145483, acc = 0.855000\n",
      "epoch [ 206] L = 39.137748, acc = 0.855000\n",
      "epoch [ 207] L = 39.130140, acc = 0.855000\n",
      "epoch [ 208] L = 39.122657, acc = 0.855000\n",
      "epoch [ 209] L = 39.115296, acc = 0.855000\n",
      "epoch [ 210] L = 39.108056, acc = 0.855000\n",
      "epoch [ 211] L = 39.100933, acc = 0.855000\n",
      "epoch [ 212] L = 39.093926, acc = 0.855000\n",
      "epoch [ 213] L = 39.087032, acc = 0.855000\n",
      "epoch [ 214] L = 39.080248, acc = 0.855000\n",
      "epoch [ 215] L = 39.073574, acc = 0.855000\n",
      "epoch [ 216] L = 39.067006, acc = 0.855000\n",
      "epoch [ 217] L = 39.060543, acc = 0.855000\n",
      "epoch [ 218] L = 39.054183, acc = 0.855000\n",
      "epoch [ 219] L = 39.047923, acc = 0.855000\n",
      "epoch [ 220] L = 39.041763, acc = 0.855000\n",
      "epoch [ 221] L = 39.035699, acc = 0.855000\n",
      "epoch [ 222] L = 39.029731, acc = 0.855000\n",
      "epoch [ 223] L = 39.023855, acc = 0.855000\n",
      "epoch [ 224] L = 39.018072, acc = 0.855000\n",
      "epoch [ 225] L = 39.012378, acc = 0.855000\n",
      "epoch [ 226] L = 39.006773, acc = 0.855000\n",
      "epoch [ 227] L = 39.001254, acc = 0.855000\n",
      "epoch [ 228] L = 38.995821, acc = 0.855000\n",
      "epoch [ 229] L = 38.990470, acc = 0.855000\n",
      "epoch [ 230] L = 38.985202, acc = 0.855000\n",
      "epoch [ 231] L = 38.980014, acc = 0.855000\n",
      "epoch [ 232] L = 38.974905, acc = 0.855000\n",
      "epoch [ 233] L = 38.969874, acc = 0.855000\n",
      "epoch [ 234] L = 38.964918, acc = 0.855000\n",
      "epoch [ 235] L = 38.960037, acc = 0.855000\n",
      "epoch [ 236] L = 38.955230, acc = 0.855000\n",
      "epoch [ 237] L = 38.950495, acc = 0.855000\n",
      "epoch [ 238] L = 38.945830, acc = 0.855000\n",
      "epoch [ 239] L = 38.941235, acc = 0.855000\n",
      "epoch [ 240] L = 38.936708, acc = 0.855000\n",
      "epoch [ 241] L = 38.932248, acc = 0.855000\n",
      "epoch [ 242] L = 38.927854, acc = 0.855000\n",
      "epoch [ 243] L = 38.923525, acc = 0.855000\n",
      "epoch [ 244] L = 38.919259, acc = 0.855000\n",
      "epoch [ 245] L = 38.915055, acc = 0.855000\n",
      "epoch [ 246] L = 38.910913, acc = 0.855000\n",
      "epoch [ 247] L = 38.906831, acc = 0.855000\n",
      "epoch [ 248] L = 38.902808, acc = 0.855000\n",
      "epoch [ 249] L = 38.898843, acc = 0.855000\n",
      "epoch [ 250] L = 38.894936, acc = 0.855000\n",
      "epoch [ 251] L = 38.891084, acc = 0.855000\n",
      "epoch [ 252] L = 38.887288, acc = 0.855000\n",
      "epoch [ 253] L = 38.883546, acc = 0.855000\n",
      "epoch [ 254] L = 38.879857, acc = 0.855000\n",
      "epoch [ 255] L = 38.876220, acc = 0.855000\n",
      "epoch [ 256] L = 38.872635, acc = 0.855000\n",
      "epoch [ 257] L = 38.869101, acc = 0.855000\n",
      "epoch [ 258] L = 38.865616, acc = 0.855000\n",
      "epoch [ 259] L = 38.862180, acc = 0.855000\n",
      "epoch [ 260] L = 38.858792, acc = 0.855000\n",
      "epoch [ 261] L = 38.855451, acc = 0.855000\n",
      "epoch [ 262] L = 38.852157, acc = 0.855000\n",
      "epoch [ 263] L = 38.848909, acc = 0.855000\n",
      "epoch [ 264] L = 38.845705, acc = 0.855000\n",
      "epoch [ 265] L = 38.842545, acc = 0.855000\n",
      "epoch [ 266] L = 38.839429, acc = 0.855000\n",
      "epoch [ 267] L = 38.836356, acc = 0.855000\n",
      "epoch [ 268] L = 38.833324, acc = 0.855000\n",
      "epoch [ 269] L = 38.830334, acc = 0.850000\n",
      "epoch [ 270] L = 38.827384, acc = 0.850000\n",
      "epoch [ 271] L = 38.824474, acc = 0.850000\n",
      "epoch [ 272] L = 38.821604, acc = 0.850000\n",
      "epoch [ 273] L = 38.818772, acc = 0.850000\n",
      "epoch [ 274] L = 38.815978, acc = 0.850000\n",
      "epoch [ 275] L = 38.813221, acc = 0.850000\n",
      "epoch [ 276] L = 38.810502, acc = 0.850000\n",
      "epoch [ 277] L = 38.807818, acc = 0.850000\n",
      "epoch [ 278] L = 38.805170, acc = 0.850000\n",
      "epoch [ 279] L = 38.802557, acc = 0.850000\n",
      "epoch [ 280] L = 38.799978, acc = 0.850000\n",
      "epoch [ 281] L = 38.797433, acc = 0.850000\n",
      "epoch [ 282] L = 38.794922, acc = 0.850000\n",
      "epoch [ 283] L = 38.792443, acc = 0.850000\n",
      "epoch [ 284] L = 38.789996, acc = 0.850000\n",
      "epoch [ 285] L = 38.787582, acc = 0.845000\n",
      "epoch [ 286] L = 38.785198, acc = 0.845000\n",
      "epoch [ 287] L = 38.782845, acc = 0.845000\n",
      "epoch [ 288] L = 38.780523, acc = 0.845000\n",
      "epoch [ 289] L = 38.778230, acc = 0.845000\n",
      "epoch [ 290] L = 38.775967, acc = 0.845000\n",
      "epoch [ 291] L = 38.773732, acc = 0.845000\n",
      "epoch [ 292] L = 38.771526, acc = 0.845000\n",
      "epoch [ 293] L = 38.769348, acc = 0.845000\n",
      "epoch [ 294] L = 38.767197, acc = 0.845000\n",
      "epoch [ 295] L = 38.765073, acc = 0.845000\n",
      "epoch [ 296] L = 38.762976, acc = 0.845000\n",
      "epoch [ 297] L = 38.760906, acc = 0.845000\n",
      "epoch [ 298] L = 38.758861, acc = 0.845000\n",
      "epoch [ 299] L = 38.756841, acc = 0.845000\n",
      "epoch [ 300] L = 38.754847, acc = 0.845000\n",
      "epoch [ 301] L = 38.752877, acc = 0.845000\n",
      "epoch [ 302] L = 38.750931, acc = 0.845000\n",
      "epoch [ 303] L = 38.749010, acc = 0.845000\n",
      "epoch [ 304] L = 38.747111, acc = 0.845000\n",
      "epoch [ 305] L = 38.745237, acc = 0.845000\n",
      "epoch [ 306] L = 38.743384, acc = 0.845000\n",
      "epoch [ 307] L = 38.741555, acc = 0.845000\n",
      "epoch [ 308] L = 38.739747, acc = 0.845000\n",
      "epoch [ 309] L = 38.737962, acc = 0.845000\n",
      "epoch [ 310] L = 38.736197, acc = 0.845000\n",
      "epoch [ 311] L = 38.734454, acc = 0.845000\n",
      "epoch [ 312] L = 38.732732, acc = 0.845000\n",
      "epoch [ 313] L = 38.731030, acc = 0.850000\n",
      "epoch [ 314] L = 38.729348, acc = 0.850000\n",
      "epoch [ 315] L = 38.727687, acc = 0.850000\n",
      "epoch [ 316] L = 38.726045, acc = 0.850000\n",
      "epoch [ 317] L = 38.724422, acc = 0.850000\n",
      "epoch [ 318] L = 38.722818, acc = 0.850000\n",
      "epoch [ 319] L = 38.721233, acc = 0.850000\n",
      "epoch [ 320] L = 38.719667, acc = 0.850000\n",
      "epoch [ 321] L = 38.718118, acc = 0.850000\n",
      "epoch [ 322] L = 38.716588, acc = 0.850000\n",
      "epoch [ 323] L = 38.715075, acc = 0.850000\n",
      "epoch [ 324] L = 38.713579, acc = 0.850000\n",
      "epoch [ 325] L = 38.712101, acc = 0.850000\n",
      "epoch [ 326] L = 38.710640, acc = 0.850000\n",
      "epoch [ 327] L = 38.709195, acc = 0.850000\n",
      "epoch [ 328] L = 38.707767, acc = 0.850000\n",
      "epoch [ 329] L = 38.706354, acc = 0.850000\n",
      "epoch [ 330] L = 38.704958, acc = 0.850000\n",
      "epoch [ 331] L = 38.703578, acc = 0.850000\n",
      "epoch [ 332] L = 38.702212, acc = 0.850000\n",
      "epoch [ 333] L = 38.700862, acc = 0.850000\n",
      "epoch [ 334] L = 38.699528, acc = 0.850000\n",
      "epoch [ 335] L = 38.698208, acc = 0.850000\n",
      "epoch [ 336] L = 38.696902, acc = 0.850000\n",
      "epoch [ 337] L = 38.695611, acc = 0.850000\n",
      "epoch [ 338] L = 38.694334, acc = 0.850000\n",
      "epoch [ 339] L = 38.693071, acc = 0.850000\n",
      "epoch [ 340] L = 38.691822, acc = 0.850000\n",
      "epoch [ 341] L = 38.690586, acc = 0.850000\n",
      "epoch [ 342] L = 38.689364, acc = 0.850000\n",
      "epoch [ 343] L = 38.688155, acc = 0.850000\n",
      "epoch [ 344] L = 38.686959, acc = 0.850000\n",
      "epoch [ 345] L = 38.685776, acc = 0.850000\n",
      "epoch [ 346] L = 38.684605, acc = 0.850000\n",
      "epoch [ 347] L = 38.683447, acc = 0.850000\n",
      "epoch [ 348] L = 38.682301, acc = 0.850000\n",
      "epoch [ 349] L = 38.681167, acc = 0.850000\n",
      "epoch [ 350] L = 38.680046, acc = 0.850000\n",
      "epoch [ 351] L = 38.678936, acc = 0.850000\n",
      "epoch [ 352] L = 38.677837, acc = 0.850000\n",
      "epoch [ 353] L = 38.676750, acc = 0.850000\n",
      "epoch [ 354] L = 38.675675, acc = 0.850000\n",
      "epoch [ 355] L = 38.674610, acc = 0.850000\n",
      "epoch [ 356] L = 38.673557, acc = 0.850000\n",
      "epoch [ 357] L = 38.672514, acc = 0.850000\n",
      "epoch [ 358] L = 38.671482, acc = 0.850000\n",
      "epoch [ 359] L = 38.670460, acc = 0.850000\n",
      "epoch [ 360] L = 38.669449, acc = 0.850000\n",
      "epoch [ 361] L = 38.668448, acc = 0.850000\n",
      "epoch [ 362] L = 38.667458, acc = 0.850000\n",
      "epoch [ 363] L = 38.666477, acc = 0.850000\n",
      "epoch [ 364] L = 38.665506, acc = 0.850000\n",
      "epoch [ 365] L = 38.664545, acc = 0.850000\n",
      "epoch [ 366] L = 38.663593, acc = 0.850000\n",
      "epoch [ 367] L = 38.662651, acc = 0.850000\n",
      "epoch [ 368] L = 38.661718, acc = 0.850000\n",
      "epoch [ 369] L = 38.660794, acc = 0.850000\n",
      "epoch [ 370] L = 38.659879, acc = 0.850000\n",
      "epoch [ 371] L = 38.658973, acc = 0.850000\n",
      "epoch [ 372] L = 38.658076, acc = 0.850000\n",
      "epoch [ 373] L = 38.657188, acc = 0.850000\n",
      "epoch [ 374] L = 38.656308, acc = 0.850000\n",
      "epoch [ 375] L = 38.655437, acc = 0.850000\n",
      "epoch [ 376] L = 38.654574, acc = 0.850000\n",
      "epoch [ 377] L = 38.653719, acc = 0.850000\n",
      "epoch [ 378] L = 38.652872, acc = 0.850000\n",
      "epoch [ 379] L = 38.652034, acc = 0.850000\n",
      "epoch [ 380] L = 38.651203, acc = 0.850000\n",
      "epoch [ 381] L = 38.650380, acc = 0.850000\n",
      "epoch [ 382] L = 38.649565, acc = 0.850000\n",
      "epoch [ 383] L = 38.648758, acc = 0.850000\n",
      "epoch [ 384] L = 38.647958, acc = 0.850000\n",
      "epoch [ 385] L = 38.647165, acc = 0.850000\n",
      "epoch [ 386] L = 38.646379, acc = 0.850000\n",
      "epoch [ 387] L = 38.645601, acc = 0.850000\n",
      "epoch [ 388] L = 38.644830, acc = 0.850000\n",
      "epoch [ 389] L = 38.644066, acc = 0.850000\n",
      "epoch [ 390] L = 38.643309, acc = 0.850000\n",
      "epoch [ 391] L = 38.642559, acc = 0.850000\n",
      "epoch [ 392] L = 38.641815, acc = 0.850000\n",
      "epoch [ 393] L = 38.641079, acc = 0.850000\n",
      "epoch [ 394] L = 38.640348, acc = 0.850000\n",
      "epoch [ 395] L = 38.639624, acc = 0.850000\n",
      "epoch [ 396] L = 38.638907, acc = 0.850000\n",
      "epoch [ 397] L = 38.638196, acc = 0.850000\n",
      "epoch [ 398] L = 38.637491, acc = 0.850000\n",
      "epoch [ 399] L = 38.636793, acc = 0.850000\n",
      "epoch [ 400] L = 38.636100, acc = 0.850000\n",
      "epoch [ 401] L = 38.635413, acc = 0.850000\n",
      "epoch [ 402] L = 38.634733, acc = 0.850000\n",
      "epoch [ 403] L = 38.634058, acc = 0.850000\n",
      "epoch [ 404] L = 38.633389, acc = 0.850000\n",
      "epoch [ 405] L = 38.632725, acc = 0.850000\n",
      "epoch [ 406] L = 38.632068, acc = 0.850000\n",
      "epoch [ 407] L = 38.631416, acc = 0.850000\n",
      "epoch [ 408] L = 38.630769, acc = 0.850000\n",
      "epoch [ 409] L = 38.630128, acc = 0.850000\n",
      "epoch [ 410] L = 38.629492, acc = 0.850000\n",
      "epoch [ 411] L = 38.628861, acc = 0.850000\n",
      "epoch [ 412] L = 38.628236, acc = 0.850000\n",
      "epoch [ 413] L = 38.627616, acc = 0.850000\n",
      "epoch [ 414] L = 38.627000, acc = 0.850000\n",
      "epoch [ 415] L = 38.626390, acc = 0.850000\n",
      "epoch [ 416] L = 38.625785, acc = 0.850000\n",
      "epoch [ 417] L = 38.625185, acc = 0.850000\n",
      "epoch [ 418] L = 38.624589, acc = 0.850000\n",
      "epoch [ 419] L = 38.623999, acc = 0.850000\n",
      "epoch [ 420] L = 38.623413, acc = 0.850000\n",
      "epoch [ 421] L = 38.622831, acc = 0.850000\n",
      "epoch [ 422] L = 38.622255, acc = 0.850000\n",
      "epoch [ 423] L = 38.621682, acc = 0.850000\n",
      "epoch [ 424] L = 38.621115, acc = 0.850000\n",
      "epoch [ 425] L = 38.620552, acc = 0.850000\n",
      "epoch [ 426] L = 38.619993, acc = 0.850000\n",
      "epoch [ 427] L = 38.619438, acc = 0.850000\n",
      "epoch [ 428] L = 38.618888, acc = 0.850000\n",
      "epoch [ 429] L = 38.618342, acc = 0.850000\n",
      "epoch [ 430] L = 38.617800, acc = 0.850000\n",
      "epoch [ 431] L = 38.617262, acc = 0.850000\n",
      "epoch [ 432] L = 38.616728, acc = 0.850000\n",
      "epoch [ 433] L = 38.616199, acc = 0.850000\n",
      "epoch [ 434] L = 38.615673, acc = 0.850000\n",
      "epoch [ 435] L = 38.615151, acc = 0.850000\n",
      "epoch [ 436] L = 38.614633, acc = 0.850000\n",
      "epoch [ 437] L = 38.614119, acc = 0.850000\n",
      "epoch [ 438] L = 38.613609, acc = 0.850000\n",
      "epoch [ 439] L = 38.613102, acc = 0.850000\n",
      "epoch [ 440] L = 38.612599, acc = 0.850000\n",
      "epoch [ 441] L = 38.612100, acc = 0.850000\n",
      "epoch [ 442] L = 38.611604, acc = 0.850000\n",
      "epoch [ 443] L = 38.611112, acc = 0.850000\n",
      "epoch [ 444] L = 38.610624, acc = 0.850000\n",
      "epoch [ 445] L = 38.610138, acc = 0.850000\n",
      "epoch [ 446] L = 38.609657, acc = 0.850000\n",
      "epoch [ 447] L = 38.609178, acc = 0.850000\n",
      "epoch [ 448] L = 38.608703, acc = 0.850000\n",
      "epoch [ 449] L = 38.608232, acc = 0.850000\n",
      "epoch [ 450] L = 38.607763, acc = 0.850000\n",
      "epoch [ 451] L = 38.607298, acc = 0.850000\n",
      "epoch [ 452] L = 38.606836, acc = 0.850000\n",
      "epoch [ 453] L = 38.606378, acc = 0.850000\n",
      "epoch [ 454] L = 38.605922, acc = 0.850000\n",
      "epoch [ 455] L = 38.605469, acc = 0.850000\n",
      "epoch [ 456] L = 38.605020, acc = 0.850000\n",
      "epoch [ 457] L = 38.604573, acc = 0.850000\n",
      "epoch [ 458] L = 38.604130, acc = 0.850000\n",
      "epoch [ 459] L = 38.603689, acc = 0.850000\n",
      "epoch [ 460] L = 38.603252, acc = 0.850000\n",
      "epoch [ 461] L = 38.602817, acc = 0.850000\n",
      "epoch [ 462] L = 38.602385, acc = 0.850000\n",
      "epoch [ 463] L = 38.601956, acc = 0.850000\n",
      "epoch [ 464] L = 38.601529, acc = 0.850000\n",
      "epoch [ 465] L = 38.601106, acc = 0.850000\n",
      "epoch [ 466] L = 38.600685, acc = 0.850000\n",
      "epoch [ 467] L = 38.600267, acc = 0.850000\n",
      "epoch [ 468] L = 38.599851, acc = 0.850000\n",
      "epoch [ 469] L = 38.599438, acc = 0.850000\n",
      "epoch [ 470] L = 38.599028, acc = 0.850000\n",
      "epoch [ 471] L = 38.598620, acc = 0.850000\n",
      "epoch [ 472] L = 38.598215, acc = 0.850000\n",
      "epoch [ 473] L = 38.597812, acc = 0.850000\n",
      "epoch [ 474] L = 38.597412, acc = 0.850000\n",
      "epoch [ 475] L = 38.597014, acc = 0.850000\n",
      "epoch [ 476] L = 38.596619, acc = 0.850000\n",
      "epoch [ 477] L = 38.596226, acc = 0.850000\n",
      "epoch [ 478] L = 38.595835, acc = 0.850000\n",
      "epoch [ 479] L = 38.595447, acc = 0.850000\n",
      "epoch [ 480] L = 38.595061, acc = 0.850000\n",
      "epoch [ 481] L = 38.594677, acc = 0.850000\n",
      "epoch [ 482] L = 38.594296, acc = 0.850000\n",
      "epoch [ 483] L = 38.593917, acc = 0.850000\n",
      "epoch [ 484] L = 38.593540, acc = 0.850000\n",
      "epoch [ 485] L = 38.593165, acc = 0.850000\n",
      "epoch [ 486] L = 38.592792, acc = 0.850000\n",
      "epoch [ 487] L = 38.592422, acc = 0.850000\n",
      "epoch [ 488] L = 38.592054, acc = 0.850000\n",
      "epoch [ 489] L = 38.591687, acc = 0.850000\n",
      "epoch [ 490] L = 38.591323, acc = 0.850000\n",
      "epoch [ 491] L = 38.590961, acc = 0.850000\n",
      "epoch [ 492] L = 38.590601, acc = 0.850000\n",
      "epoch [ 493] L = 38.590243, acc = 0.850000\n",
      "epoch [ 494] L = 38.589887, acc = 0.850000\n",
      "epoch [ 495] L = 38.589533, acc = 0.850000\n",
      "epoch [ 496] L = 38.589181, acc = 0.850000\n",
      "epoch [ 497] L = 38.588830, acc = 0.850000\n",
      "epoch [ 498] L = 38.588482, acc = 0.850000\n",
      "epoch [ 499] L = 38.588136, acc = 0.850000\n",
      "epoch [ 500] L = 38.587791, acc = 0.850000\n",
      "epoch [ 501] L = 38.587448, acc = 0.850000\n",
      "epoch [ 502] L = 38.587107, acc = 0.850000\n",
      "epoch [ 503] L = 38.586768, acc = 0.850000\n",
      "epoch [ 504] L = 38.586431, acc = 0.850000\n",
      "epoch [ 505] L = 38.586095, acc = 0.850000\n",
      "epoch [ 506] L = 38.585761, acc = 0.850000\n",
      "epoch [ 507] L = 38.585429, acc = 0.850000\n",
      "epoch [ 508] L = 38.585099, acc = 0.850000\n",
      "epoch [ 509] L = 38.584770, acc = 0.850000\n",
      "epoch [ 510] L = 38.584443, acc = 0.850000\n",
      "epoch [ 511] L = 38.584118, acc = 0.850000\n",
      "epoch [ 512] L = 38.583794, acc = 0.850000\n",
      "epoch [ 513] L = 38.583472, acc = 0.850000\n",
      "epoch [ 514] L = 38.583151, acc = 0.850000\n",
      "epoch [ 515] L = 38.582832, acc = 0.850000\n",
      "epoch [ 516] L = 38.582515, acc = 0.850000\n",
      "epoch [ 517] L = 38.582199, acc = 0.850000\n",
      "epoch [ 518] L = 38.581885, acc = 0.850000\n",
      "epoch [ 519] L = 38.581572, acc = 0.850000\n",
      "epoch [ 520] L = 38.581261, acc = 0.850000\n",
      "epoch [ 521] L = 38.580951, acc = 0.850000\n",
      "epoch [ 522] L = 38.580643, acc = 0.850000\n",
      "epoch [ 523] L = 38.580336, acc = 0.850000\n",
      "epoch [ 524] L = 38.580030, acc = 0.850000\n",
      "epoch [ 525] L = 38.579726, acc = 0.850000\n",
      "epoch [ 526] L = 38.579424, acc = 0.850000\n",
      "epoch [ 527] L = 38.579123, acc = 0.850000\n",
      "epoch [ 528] L = 38.578823, acc = 0.850000\n",
      "epoch [ 529] L = 38.578525, acc = 0.850000\n",
      "epoch [ 530] L = 38.578228, acc = 0.850000\n",
      "epoch [ 531] L = 38.577932, acc = 0.850000\n",
      "epoch [ 532] L = 38.577638, acc = 0.850000\n",
      "epoch [ 533] L = 38.577345, acc = 0.850000\n",
      "epoch [ 534] L = 38.577053, acc = 0.850000\n",
      "epoch [ 535] L = 38.576763, acc = 0.850000\n",
      "epoch [ 536] L = 38.576474, acc = 0.850000\n",
      "epoch [ 537] L = 38.576186, acc = 0.850000\n",
      "epoch [ 538] L = 38.575899, acc = 0.850000\n",
      "epoch [ 539] L = 38.575614, acc = 0.850000\n",
      "epoch [ 540] L = 38.575330, acc = 0.850000\n",
      "epoch [ 541] L = 38.575047, acc = 0.850000\n",
      "epoch [ 542] L = 38.574765, acc = 0.850000\n",
      "epoch [ 543] L = 38.574485, acc = 0.850000\n",
      "epoch [ 544] L = 38.574205, acc = 0.850000\n",
      "epoch [ 545] L = 38.573927, acc = 0.850000\n",
      "epoch [ 546] L = 38.573650, acc = 0.850000\n",
      "epoch [ 547] L = 38.573375, acc = 0.850000\n",
      "epoch [ 548] L = 38.573100, acc = 0.850000\n",
      "epoch [ 549] L = 38.572826, acc = 0.850000\n",
      "epoch [ 550] L = 38.572554, acc = 0.850000\n",
      "epoch [ 551] L = 38.572283, acc = 0.850000\n",
      "epoch [ 552] L = 38.572013, acc = 0.850000\n",
      "epoch [ 553] L = 38.571743, acc = 0.850000\n",
      "epoch [ 554] L = 38.571475, acc = 0.850000\n",
      "epoch [ 555] L = 38.571209, acc = 0.850000\n",
      "epoch [ 556] L = 38.570943, acc = 0.850000\n",
      "epoch [ 557] L = 38.570678, acc = 0.850000\n",
      "epoch [ 558] L = 38.570414, acc = 0.850000\n",
      "epoch [ 559] L = 38.570151, acc = 0.850000\n",
      "epoch [ 560] L = 38.569889, acc = 0.850000\n",
      "epoch [ 561] L = 38.569629, acc = 0.850000\n",
      "epoch [ 562] L = 38.569369, acc = 0.850000\n",
      "epoch [ 563] L = 38.569110, acc = 0.850000\n",
      "epoch [ 564] L = 38.568852, acc = 0.850000\n",
      "epoch [ 565] L = 38.568596, acc = 0.850000\n",
      "epoch [ 566] L = 38.568340, acc = 0.850000\n",
      "epoch [ 567] L = 38.568085, acc = 0.850000\n",
      "epoch [ 568] L = 38.567831, acc = 0.850000\n",
      "epoch [ 569] L = 38.567578, acc = 0.850000\n",
      "epoch [ 570] L = 38.567326, acc = 0.850000\n",
      "epoch [ 571] L = 38.567075, acc = 0.850000\n",
      "epoch [ 572] L = 38.566825, acc = 0.850000\n",
      "epoch [ 573] L = 38.566575, acc = 0.850000\n",
      "epoch [ 574] L = 38.566327, acc = 0.850000\n",
      "epoch [ 575] L = 38.566079, acc = 0.850000\n",
      "epoch [ 576] L = 38.565833, acc = 0.850000\n",
      "epoch [ 577] L = 38.565587, acc = 0.850000\n",
      "epoch [ 578] L = 38.565342, acc = 0.850000\n",
      "epoch [ 579] L = 38.565098, acc = 0.850000\n",
      "epoch [ 580] L = 38.564855, acc = 0.850000\n",
      "epoch [ 581] L = 38.564613, acc = 0.850000\n",
      "epoch [ 582] L = 38.564371, acc = 0.850000\n",
      "epoch [ 583] L = 38.564130, acc = 0.850000\n",
      "epoch [ 584] L = 38.563891, acc = 0.850000\n",
      "epoch [ 585] L = 38.563652, acc = 0.850000\n",
      "epoch [ 586] L = 38.563413, acc = 0.850000\n",
      "epoch [ 587] L = 38.563176, acc = 0.850000\n",
      "epoch [ 588] L = 38.562939, acc = 0.850000\n",
      "epoch [ 589] L = 38.562703, acc = 0.850000\n",
      "epoch [ 590] L = 38.562468, acc = 0.850000\n",
      "epoch [ 591] L = 38.562234, acc = 0.850000\n",
      "epoch [ 592] L = 38.562001, acc = 0.850000\n",
      "epoch [ 593] L = 38.561768, acc = 0.850000\n",
      "epoch [ 594] L = 38.561536, acc = 0.850000\n",
      "epoch [ 595] L = 38.561304, acc = 0.850000\n",
      "epoch [ 596] L = 38.561074, acc = 0.850000\n",
      "epoch [ 597] L = 38.560844, acc = 0.850000\n",
      "epoch [ 598] L = 38.560615, acc = 0.850000\n",
      "epoch [ 599] L = 38.560387, acc = 0.850000\n",
      "epoch [ 600] L = 38.560159, acc = 0.850000\n",
      "epoch [ 601] L = 38.559932, acc = 0.850000\n",
      "epoch [ 602] L = 38.559706, acc = 0.850000\n",
      "epoch [ 603] L = 38.559481, acc = 0.850000\n",
      "epoch [ 604] L = 38.559256, acc = 0.850000\n",
      "epoch [ 605] L = 38.559032, acc = 0.850000\n",
      "epoch [ 606] L = 38.558808, acc = 0.850000\n",
      "epoch [ 607] L = 38.558586, acc = 0.850000\n",
      "epoch [ 608] L = 38.558364, acc = 0.850000\n",
      "epoch [ 609] L = 38.558142, acc = 0.850000\n",
      "epoch [ 610] L = 38.557922, acc = 0.850000\n",
      "epoch [ 611] L = 38.557701, acc = 0.850000\n",
      "epoch [ 612] L = 38.557482, acc = 0.850000\n",
      "epoch [ 613] L = 38.557263, acc = 0.850000\n",
      "epoch [ 614] L = 38.557045, acc = 0.850000\n",
      "epoch [ 615] L = 38.556828, acc = 0.850000\n",
      "epoch [ 616] L = 38.556611, acc = 0.850000\n",
      "epoch [ 617] L = 38.556395, acc = 0.850000\n",
      "epoch [ 618] L = 38.556179, acc = 0.850000\n",
      "epoch [ 619] L = 38.555964, acc = 0.850000\n",
      "epoch [ 620] L = 38.555750, acc = 0.850000\n",
      "epoch [ 621] L = 38.555536, acc = 0.850000\n",
      "epoch [ 622] L = 38.555323, acc = 0.850000\n",
      "epoch [ 623] L = 38.555110, acc = 0.850000\n",
      "epoch [ 624] L = 38.554898, acc = 0.850000\n",
      "epoch [ 625] L = 38.554687, acc = 0.850000\n",
      "epoch [ 626] L = 38.554476, acc = 0.850000\n",
      "epoch [ 627] L = 38.554266, acc = 0.850000\n",
      "epoch [ 628] L = 38.554056, acc = 0.850000\n",
      "epoch [ 629] L = 38.553847, acc = 0.850000\n",
      "epoch [ 630] L = 38.553639, acc = 0.850000\n",
      "epoch [ 631] L = 38.553431, acc = 0.850000\n",
      "epoch [ 632] L = 38.553223, acc = 0.850000\n",
      "epoch [ 633] L = 38.553017, acc = 0.850000\n",
      "epoch [ 634] L = 38.552810, acc = 0.850000\n",
      "epoch [ 635] L = 38.552605, acc = 0.850000\n",
      "epoch [ 636] L = 38.552400, acc = 0.850000\n",
      "epoch [ 637] L = 38.552195, acc = 0.850000\n",
      "epoch [ 638] L = 38.551991, acc = 0.850000\n",
      "epoch [ 639] L = 38.551787, acc = 0.850000\n",
      "epoch [ 640] L = 38.551584, acc = 0.850000\n",
      "epoch [ 641] L = 38.551382, acc = 0.850000\n",
      "epoch [ 642] L = 38.551180, acc = 0.850000\n",
      "epoch [ 643] L = 38.550978, acc = 0.850000\n",
      "epoch [ 644] L = 38.550777, acc = 0.850000\n",
      "epoch [ 645] L = 38.550577, acc = 0.850000\n",
      "epoch [ 646] L = 38.550377, acc = 0.850000\n",
      "epoch [ 647] L = 38.550178, acc = 0.850000\n",
      "epoch [ 648] L = 38.549979, acc = 0.850000\n",
      "epoch [ 649] L = 38.549780, acc = 0.850000\n",
      "epoch [ 650] L = 38.549582, acc = 0.850000\n",
      "epoch [ 651] L = 38.549385, acc = 0.850000\n",
      "epoch [ 652] L = 38.549188, acc = 0.850000\n",
      "epoch [ 653] L = 38.548992, acc = 0.850000\n",
      "epoch [ 654] L = 38.548796, acc = 0.850000\n",
      "epoch [ 655] L = 38.548600, acc = 0.850000\n",
      "epoch [ 656] L = 38.548405, acc = 0.850000\n",
      "epoch [ 657] L = 38.548211, acc = 0.850000\n",
      "epoch [ 658] L = 38.548016, acc = 0.850000\n",
      "epoch [ 659] L = 38.547823, acc = 0.850000\n",
      "epoch [ 660] L = 38.547630, acc = 0.850000\n",
      "epoch [ 661] L = 38.547437, acc = 0.850000\n",
      "epoch [ 662] L = 38.547245, acc = 0.850000\n",
      "epoch [ 663] L = 38.547053, acc = 0.850000\n",
      "epoch [ 664] L = 38.546861, acc = 0.850000\n",
      "epoch [ 665] L = 38.546671, acc = 0.850000\n",
      "epoch [ 666] L = 38.546480, acc = 0.850000\n",
      "epoch [ 667] L = 38.546290, acc = 0.850000\n",
      "epoch [ 668] L = 38.546101, acc = 0.850000\n",
      "epoch [ 669] L = 38.545911, acc = 0.850000\n",
      "epoch [ 670] L = 38.545723, acc = 0.850000\n",
      "epoch [ 671] L = 38.545534, acc = 0.850000\n",
      "epoch [ 672] L = 38.545346, acc = 0.850000\n",
      "epoch [ 673] L = 38.545159, acc = 0.850000\n",
      "epoch [ 674] L = 38.544972, acc = 0.850000\n",
      "epoch [ 675] L = 38.544785, acc = 0.850000\n",
      "epoch [ 676] L = 38.544599, acc = 0.850000\n",
      "epoch [ 677] L = 38.544414, acc = 0.850000\n",
      "epoch [ 678] L = 38.544228, acc = 0.850000\n",
      "epoch [ 679] L = 38.544043, acc = 0.850000\n",
      "epoch [ 680] L = 38.543859, acc = 0.850000\n",
      "epoch [ 681] L = 38.543675, acc = 0.850000\n",
      "epoch [ 682] L = 38.543491, acc = 0.850000\n",
      "epoch [ 683] L = 38.543307, acc = 0.850000\n",
      "epoch [ 684] L = 38.543125, acc = 0.850000\n",
      "epoch [ 685] L = 38.542942, acc = 0.850000\n",
      "epoch [ 686] L = 38.542760, acc = 0.850000\n",
      "epoch [ 687] L = 38.542578, acc = 0.850000\n",
      "epoch [ 688] L = 38.542397, acc = 0.850000\n",
      "epoch [ 689] L = 38.542216, acc = 0.850000\n",
      "epoch [ 690] L = 38.542035, acc = 0.850000\n",
      "epoch [ 691] L = 38.541855, acc = 0.850000\n",
      "epoch [ 692] L = 38.541675, acc = 0.850000\n",
      "epoch [ 693] L = 38.541496, acc = 0.850000\n",
      "epoch [ 694] L = 38.541316, acc = 0.850000\n",
      "epoch [ 695] L = 38.541138, acc = 0.850000\n",
      "epoch [ 696] L = 38.540959, acc = 0.850000\n",
      "epoch [ 697] L = 38.540781, acc = 0.850000\n",
      "epoch [ 698] L = 38.540604, acc = 0.850000\n",
      "epoch [ 699] L = 38.540427, acc = 0.850000\n",
      "epoch [ 700] L = 38.540250, acc = 0.850000\n",
      "epoch [ 701] L = 38.540073, acc = 0.850000\n",
      "epoch [ 702] L = 38.539897, acc = 0.850000\n",
      "epoch [ 703] L = 38.539721, acc = 0.850000\n",
      "epoch [ 704] L = 38.539546, acc = 0.850000\n",
      "epoch [ 705] L = 38.539371, acc = 0.850000\n",
      "epoch [ 706] L = 38.539196, acc = 0.850000\n",
      "epoch [ 707] L = 38.539021, acc = 0.850000\n",
      "epoch [ 708] L = 38.538847, acc = 0.850000\n",
      "epoch [ 709] L = 38.538674, acc = 0.850000\n",
      "epoch [ 710] L = 38.538500, acc = 0.850000\n",
      "epoch [ 711] L = 38.538327, acc = 0.850000\n",
      "epoch [ 712] L = 38.538154, acc = 0.850000\n",
      "epoch [ 713] L = 38.537982, acc = 0.850000\n",
      "epoch [ 714] L = 38.537810, acc = 0.850000\n",
      "epoch [ 715] L = 38.537638, acc = 0.850000\n",
      "epoch [ 716] L = 38.537467, acc = 0.850000\n",
      "epoch [ 717] L = 38.537296, acc = 0.850000\n",
      "epoch [ 718] L = 38.537125, acc = 0.850000\n",
      "epoch [ 719] L = 38.536955, acc = 0.850000\n",
      "epoch [ 720] L = 38.536785, acc = 0.850000\n",
      "epoch [ 721] L = 38.536615, acc = 0.850000\n",
      "epoch [ 722] L = 38.536446, acc = 0.850000\n",
      "epoch [ 723] L = 38.536277, acc = 0.850000\n",
      "epoch [ 724] L = 38.536108, acc = 0.850000\n",
      "epoch [ 725] L = 38.535939, acc = 0.850000\n",
      "epoch [ 726] L = 38.535771, acc = 0.850000\n",
      "epoch [ 727] L = 38.535603, acc = 0.850000\n",
      "epoch [ 728] L = 38.535436, acc = 0.850000\n",
      "epoch [ 729] L = 38.535269, acc = 0.850000\n",
      "epoch [ 730] L = 38.535102, acc = 0.850000\n",
      "epoch [ 731] L = 38.534935, acc = 0.850000\n",
      "epoch [ 732] L = 38.534769, acc = 0.850000\n",
      "epoch [ 733] L = 38.534603, acc = 0.850000\n",
      "epoch [ 734] L = 38.534437, acc = 0.850000\n",
      "epoch [ 735] L = 38.534272, acc = 0.850000\n",
      "epoch [ 736] L = 38.534107, acc = 0.850000\n",
      "epoch [ 737] L = 38.533942, acc = 0.850000\n",
      "epoch [ 738] L = 38.533777, acc = 0.850000\n",
      "epoch [ 739] L = 38.533613, acc = 0.850000\n",
      "epoch [ 740] L = 38.533449, acc = 0.850000\n",
      "epoch [ 741] L = 38.533286, acc = 0.850000\n",
      "epoch [ 742] L = 38.533122, acc = 0.850000\n",
      "epoch [ 743] L = 38.532959, acc = 0.850000\n",
      "epoch [ 744] L = 38.532797, acc = 0.845000\n",
      "epoch [ 745] L = 38.532634, acc = 0.845000\n",
      "epoch [ 746] L = 38.532472, acc = 0.845000\n",
      "epoch [ 747] L = 38.532310, acc = 0.845000\n",
      "epoch [ 748] L = 38.532149, acc = 0.845000\n",
      "epoch [ 749] L = 38.531987, acc = 0.845000\n",
      "epoch [ 750] L = 38.531826, acc = 0.845000\n",
      "epoch [ 751] L = 38.531666, acc = 0.845000\n",
      "epoch [ 752] L = 38.531505, acc = 0.845000\n",
      "epoch [ 753] L = 38.531345, acc = 0.845000\n",
      "epoch [ 754] L = 38.531185, acc = 0.845000\n",
      "epoch [ 755] L = 38.531025, acc = 0.845000\n",
      "epoch [ 756] L = 38.530866, acc = 0.845000\n",
      "epoch [ 757] L = 38.530707, acc = 0.845000\n",
      "epoch [ 758] L = 38.530548, acc = 0.845000\n",
      "epoch [ 759] L = 38.530390, acc = 0.845000\n",
      "epoch [ 760] L = 38.530231, acc = 0.845000\n",
      "epoch [ 761] L = 38.530073, acc = 0.845000\n",
      "epoch [ 762] L = 38.529916, acc = 0.845000\n",
      "epoch [ 763] L = 38.529758, acc = 0.845000\n",
      "epoch [ 764] L = 38.529601, acc = 0.845000\n",
      "epoch [ 765] L = 38.529444, acc = 0.845000\n",
      "epoch [ 766] L = 38.529287, acc = 0.845000\n",
      "epoch [ 767] L = 38.529131, acc = 0.845000\n",
      "epoch [ 768] L = 38.528975, acc = 0.845000\n",
      "epoch [ 769] L = 38.528819, acc = 0.845000\n",
      "epoch [ 770] L = 38.528663, acc = 0.845000\n",
      "epoch [ 771] L = 38.528508, acc = 0.845000\n",
      "epoch [ 772] L = 38.528352, acc = 0.845000\n",
      "epoch [ 773] L = 38.528198, acc = 0.845000\n",
      "epoch [ 774] L = 38.528043, acc = 0.845000\n",
      "epoch [ 775] L = 38.527889, acc = 0.845000\n",
      "epoch [ 776] L = 38.527734, acc = 0.845000\n",
      "epoch [ 777] L = 38.527580, acc = 0.845000\n",
      "epoch [ 778] L = 38.527427, acc = 0.845000\n",
      "epoch [ 779] L = 38.527273, acc = 0.845000\n",
      "epoch [ 780] L = 38.527120, acc = 0.845000\n",
      "epoch [ 781] L = 38.526967, acc = 0.845000\n",
      "epoch [ 782] L = 38.526815, acc = 0.845000\n",
      "epoch [ 783] L = 38.526662, acc = 0.845000\n",
      "epoch [ 784] L = 38.526510, acc = 0.845000\n",
      "epoch [ 785] L = 38.526358, acc = 0.845000\n",
      "epoch [ 786] L = 38.526207, acc = 0.845000\n",
      "epoch [ 787] L = 38.526055, acc = 0.845000\n",
      "epoch [ 788] L = 38.525904, acc = 0.845000\n",
      "epoch [ 789] L = 38.525753, acc = 0.845000\n",
      "epoch [ 790] L = 38.525602, acc = 0.845000\n",
      "epoch [ 791] L = 38.525452, acc = 0.845000\n",
      "epoch [ 792] L = 38.525302, acc = 0.845000\n",
      "epoch [ 793] L = 38.525152, acc = 0.845000\n",
      "epoch [ 794] L = 38.525002, acc = 0.845000\n",
      "epoch [ 795] L = 38.524852, acc = 0.845000\n",
      "epoch [ 796] L = 38.524703, acc = 0.845000\n",
      "epoch [ 797] L = 38.524554, acc = 0.845000\n",
      "epoch [ 798] L = 38.524405, acc = 0.845000\n",
      "epoch [ 799] L = 38.524256, acc = 0.845000\n",
      "epoch [ 800] L = 38.524108, acc = 0.845000\n",
      "epoch [ 801] L = 38.523960, acc = 0.845000\n",
      "epoch [ 802] L = 38.523812, acc = 0.845000\n",
      "epoch [ 803] L = 38.523664, acc = 0.845000\n",
      "epoch [ 804] L = 38.523517, acc = 0.845000\n",
      "epoch [ 805] L = 38.523369, acc = 0.845000\n",
      "epoch [ 806] L = 38.523222, acc = 0.845000\n",
      "epoch [ 807] L = 38.523075, acc = 0.845000\n",
      "epoch [ 808] L = 38.522929, acc = 0.845000\n",
      "epoch [ 809] L = 38.522782, acc = 0.845000\n",
      "epoch [ 810] L = 38.522636, acc = 0.845000\n",
      "epoch [ 811] L = 38.522490, acc = 0.845000\n",
      "epoch [ 812] L = 38.522345, acc = 0.845000\n",
      "epoch [ 813] L = 38.522199, acc = 0.845000\n",
      "epoch [ 814] L = 38.522054, acc = 0.845000\n",
      "epoch [ 815] L = 38.521909, acc = 0.845000\n",
      "epoch [ 816] L = 38.521764, acc = 0.845000\n",
      "epoch [ 817] L = 38.521619, acc = 0.845000\n",
      "epoch [ 818] L = 38.521475, acc = 0.845000\n",
      "epoch [ 819] L = 38.521331, acc = 0.845000\n",
      "epoch [ 820] L = 38.521187, acc = 0.845000\n",
      "epoch [ 821] L = 38.521043, acc = 0.845000\n",
      "epoch [ 822] L = 38.520899, acc = 0.845000\n",
      "epoch [ 823] L = 38.520756, acc = 0.845000\n",
      "epoch [ 824] L = 38.520613, acc = 0.845000\n",
      "epoch [ 825] L = 38.520470, acc = 0.845000\n",
      "epoch [ 826] L = 38.520327, acc = 0.845000\n",
      "epoch [ 827] L = 38.520185, acc = 0.845000\n",
      "epoch [ 828] L = 38.520042, acc = 0.845000\n",
      "epoch [ 829] L = 38.519900, acc = 0.845000\n",
      "epoch [ 830] L = 38.519758, acc = 0.845000\n",
      "epoch [ 831] L = 38.519617, acc = 0.845000\n",
      "epoch [ 832] L = 38.519475, acc = 0.845000\n",
      "epoch [ 833] L = 38.519334, acc = 0.845000\n",
      "epoch [ 834] L = 38.519193, acc = 0.845000\n",
      "epoch [ 835] L = 38.519052, acc = 0.845000\n",
      "epoch [ 836] L = 38.518911, acc = 0.845000\n",
      "epoch [ 837] L = 38.518771, acc = 0.845000\n",
      "epoch [ 838] L = 38.518630, acc = 0.845000\n",
      "epoch [ 839] L = 38.518490, acc = 0.845000\n",
      "epoch [ 840] L = 38.518350, acc = 0.845000\n",
      "epoch [ 841] L = 38.518211, acc = 0.845000\n",
      "epoch [ 842] L = 38.518071, acc = 0.845000\n",
      "epoch [ 843] L = 38.517932, acc = 0.845000\n",
      "epoch [ 844] L = 38.517793, acc = 0.845000\n",
      "epoch [ 845] L = 38.517654, acc = 0.845000\n",
      "epoch [ 846] L = 38.517515, acc = 0.845000\n",
      "epoch [ 847] L = 38.517377, acc = 0.845000\n",
      "epoch [ 848] L = 38.517238, acc = 0.845000\n",
      "epoch [ 849] L = 38.517100, acc = 0.845000\n",
      "epoch [ 850] L = 38.516962, acc = 0.845000\n",
      "epoch [ 851] L = 38.516825, acc = 0.845000\n",
      "epoch [ 852] L = 38.516687, acc = 0.845000\n",
      "epoch [ 853] L = 38.516550, acc = 0.845000\n",
      "epoch [ 854] L = 38.516412, acc = 0.845000\n",
      "epoch [ 855] L = 38.516275, acc = 0.845000\n",
      "epoch [ 856] L = 38.516139, acc = 0.845000\n",
      "epoch [ 857] L = 38.516002, acc = 0.845000\n",
      "epoch [ 858] L = 38.515866, acc = 0.845000\n",
      "epoch [ 859] L = 38.515729, acc = 0.845000\n",
      "epoch [ 860] L = 38.515593, acc = 0.845000\n",
      "epoch [ 861] L = 38.515458, acc = 0.845000\n",
      "epoch [ 862] L = 38.515322, acc = 0.845000\n",
      "epoch [ 863] L = 38.515186, acc = 0.845000\n",
      "epoch [ 864] L = 38.515051, acc = 0.845000\n",
      "epoch [ 865] L = 38.514916, acc = 0.845000\n",
      "epoch [ 866] L = 38.514781, acc = 0.845000\n",
      "epoch [ 867] L = 38.514646, acc = 0.845000\n",
      "epoch [ 868] L = 38.514512, acc = 0.845000\n",
      "epoch [ 869] L = 38.514377, acc = 0.845000\n",
      "epoch [ 870] L = 38.514243, acc = 0.845000\n",
      "epoch [ 871] L = 38.514109, acc = 0.845000\n",
      "epoch [ 872] L = 38.513975, acc = 0.845000\n",
      "epoch [ 873] L = 38.513842, acc = 0.845000\n",
      "epoch [ 874] L = 38.513708, acc = 0.845000\n",
      "epoch [ 875] L = 38.513575, acc = 0.845000\n",
      "epoch [ 876] L = 38.513442, acc = 0.845000\n",
      "epoch [ 877] L = 38.513309, acc = 0.845000\n",
      "epoch [ 878] L = 38.513176, acc = 0.845000\n",
      "epoch [ 879] L = 38.513043, acc = 0.845000\n",
      "epoch [ 880] L = 38.512911, acc = 0.845000\n",
      "epoch [ 881] L = 38.512779, acc = 0.845000\n",
      "epoch [ 882] L = 38.512647, acc = 0.845000\n",
      "epoch [ 883] L = 38.512515, acc = 0.845000\n",
      "epoch [ 884] L = 38.512383, acc = 0.845000\n",
      "epoch [ 885] L = 38.512251, acc = 0.845000\n",
      "epoch [ 886] L = 38.512120, acc = 0.845000\n",
      "epoch [ 887] L = 38.511989, acc = 0.845000\n",
      "epoch [ 888] L = 38.511858, acc = 0.845000\n",
      "epoch [ 889] L = 38.511727, acc = 0.845000\n",
      "epoch [ 890] L = 38.511596, acc = 0.845000\n",
      "epoch [ 891] L = 38.511466, acc = 0.845000\n",
      "epoch [ 892] L = 38.511335, acc = 0.845000\n",
      "epoch [ 893] L = 38.511205, acc = 0.845000\n",
      "epoch [ 894] L = 38.511075, acc = 0.845000\n",
      "epoch [ 895] L = 38.510945, acc = 0.845000\n",
      "epoch [ 896] L = 38.510816, acc = 0.845000\n",
      "epoch [ 897] L = 38.510686, acc = 0.845000\n",
      "epoch [ 898] L = 38.510557, acc = 0.845000\n",
      "epoch [ 899] L = 38.510428, acc = 0.845000\n",
      "epoch [ 900] L = 38.510299, acc = 0.845000\n",
      "epoch [ 901] L = 38.510170, acc = 0.845000\n",
      "epoch [ 902] L = 38.510041, acc = 0.845000\n",
      "epoch [ 903] L = 38.509913, acc = 0.845000\n",
      "epoch [ 904] L = 38.509784, acc = 0.845000\n",
      "epoch [ 905] L = 38.509656, acc = 0.845000\n",
      "epoch [ 906] L = 38.509528, acc = 0.845000\n",
      "epoch [ 907] L = 38.509400, acc = 0.845000\n",
      "epoch [ 908] L = 38.509273, acc = 0.845000\n",
      "epoch [ 909] L = 38.509145, acc = 0.845000\n",
      "epoch [ 910] L = 38.509018, acc = 0.845000\n",
      "epoch [ 911] L = 38.508890, acc = 0.845000\n",
      "epoch [ 912] L = 38.508763, acc = 0.845000\n",
      "epoch [ 913] L = 38.508637, acc = 0.845000\n",
      "epoch [ 914] L = 38.508510, acc = 0.845000\n",
      "epoch [ 915] L = 38.508383, acc = 0.845000\n",
      "epoch [ 916] L = 38.508257, acc = 0.845000\n",
      "epoch [ 917] L = 38.508131, acc = 0.845000\n",
      "epoch [ 918] L = 38.508004, acc = 0.845000\n",
      "epoch [ 919] L = 38.507879, acc = 0.845000\n",
      "epoch [ 920] L = 38.507753, acc = 0.845000\n",
      "epoch [ 921] L = 38.507627, acc = 0.845000\n",
      "epoch [ 922] L = 38.507502, acc = 0.845000\n",
      "epoch [ 923] L = 38.507376, acc = 0.845000\n",
      "epoch [ 924] L = 38.507251, acc = 0.845000\n",
      "epoch [ 925] L = 38.507126, acc = 0.845000\n",
      "epoch [ 926] L = 38.507001, acc = 0.845000\n",
      "epoch [ 927] L = 38.506877, acc = 0.845000\n",
      "epoch [ 928] L = 38.506752, acc = 0.845000\n",
      "epoch [ 929] L = 38.506628, acc = 0.845000\n",
      "epoch [ 930] L = 38.506504, acc = 0.845000\n",
      "epoch [ 931] L = 38.506380, acc = 0.845000\n",
      "epoch [ 932] L = 38.506256, acc = 0.845000\n",
      "epoch [ 933] L = 38.506132, acc = 0.845000\n",
      "epoch [ 934] L = 38.506008, acc = 0.845000\n",
      "epoch [ 935] L = 38.505885, acc = 0.845000\n",
      "epoch [ 936] L = 38.505761, acc = 0.845000\n",
      "epoch [ 937] L = 38.505638, acc = 0.845000\n",
      "epoch [ 938] L = 38.505515, acc = 0.845000\n",
      "epoch [ 939] L = 38.505392, acc = 0.845000\n",
      "epoch [ 940] L = 38.505270, acc = 0.845000\n",
      "epoch [ 941] L = 38.505147, acc = 0.845000\n",
      "epoch [ 942] L = 38.505025, acc = 0.845000\n",
      "epoch [ 943] L = 38.504903, acc = 0.845000\n",
      "epoch [ 944] L = 38.504780, acc = 0.845000\n",
      "epoch [ 945] L = 38.504658, acc = 0.845000\n",
      "epoch [ 946] L = 38.504537, acc = 0.845000\n",
      "epoch [ 947] L = 38.504415, acc = 0.845000\n",
      "epoch [ 948] L = 38.504293, acc = 0.845000\n",
      "epoch [ 949] L = 38.504172, acc = 0.845000\n",
      "epoch [ 950] L = 38.504051, acc = 0.845000\n",
      "epoch [ 951] L = 38.503930, acc = 0.845000\n",
      "epoch [ 952] L = 38.503809, acc = 0.845000\n",
      "epoch [ 953] L = 38.503688, acc = 0.845000\n",
      "epoch [ 954] L = 38.503567, acc = 0.845000\n",
      "epoch [ 955] L = 38.503447, acc = 0.845000\n",
      "epoch [ 956] L = 38.503326, acc = 0.845000\n",
      "epoch [ 957] L = 38.503206, acc = 0.845000\n",
      "epoch [ 958] L = 38.503086, acc = 0.845000\n",
      "epoch [ 959] L = 38.502966, acc = 0.845000\n",
      "epoch [ 960] L = 38.502846, acc = 0.845000\n",
      "epoch [ 961] L = 38.502727, acc = 0.845000\n",
      "epoch [ 962] L = 38.502607, acc = 0.845000\n",
      "epoch [ 963] L = 38.502488, acc = 0.845000\n",
      "epoch [ 964] L = 38.502369, acc = 0.845000\n",
      "epoch [ 965] L = 38.502250, acc = 0.845000\n",
      "epoch [ 966] L = 38.502131, acc = 0.845000\n",
      "epoch [ 967] L = 38.502012, acc = 0.845000\n",
      "epoch [ 968] L = 38.501893, acc = 0.845000\n",
      "epoch [ 969] L = 38.501775, acc = 0.845000\n",
      "epoch [ 970] L = 38.501656, acc = 0.845000\n",
      "epoch [ 971] L = 38.501538, acc = 0.845000\n",
      "epoch [ 972] L = 38.501420, acc = 0.845000\n",
      "epoch [ 973] L = 38.501302, acc = 0.845000\n",
      "epoch [ 974] L = 38.501184, acc = 0.845000\n",
      "epoch [ 975] L = 38.501066, acc = 0.845000\n",
      "epoch [ 976] L = 38.500949, acc = 0.845000\n",
      "epoch [ 977] L = 38.500831, acc = 0.845000\n",
      "epoch [ 978] L = 38.500714, acc = 0.845000\n",
      "epoch [ 979] L = 38.500597, acc = 0.845000\n",
      "epoch [ 980] L = 38.500480, acc = 0.845000\n",
      "epoch [ 981] L = 38.500363, acc = 0.845000\n",
      "epoch [ 982] L = 38.500246, acc = 0.845000\n",
      "epoch [ 983] L = 38.500130, acc = 0.845000\n",
      "epoch [ 984] L = 38.500013, acc = 0.845000\n",
      "epoch [ 985] L = 38.499897, acc = 0.845000\n",
      "epoch [ 986] L = 38.499781, acc = 0.845000\n",
      "epoch [ 987] L = 38.499665, acc = 0.845000\n",
      "epoch [ 988] L = 38.499549, acc = 0.845000\n",
      "epoch [ 989] L = 38.499433, acc = 0.845000\n",
      "epoch [ 990] L = 38.499317, acc = 0.845000\n",
      "epoch [ 991] L = 38.499202, acc = 0.845000\n",
      "epoch [ 992] L = 38.499086, acc = 0.845000\n",
      "epoch [ 993] L = 38.498971, acc = 0.845000\n",
      "epoch [ 994] L = 38.498856, acc = 0.845000\n",
      "epoch [ 995] L = 38.498741, acc = 0.845000\n",
      "epoch [ 996] L = 38.498626, acc = 0.845000\n",
      "epoch [ 997] L = 38.498511, acc = 0.845000\n",
      "epoch [ 998] L = 38.498396, acc = 0.845000\n",
      "epoch [ 999] L = 38.498282, acc = 0.845000\n",
      "epoch [1000] L = 38.498167, acc = 0.845000\n",
      "epoch [1001] L = 38.498053, acc = 0.845000\n",
      "epoch [1002] L = 38.497939, acc = 0.845000\n",
      "epoch [1003] L = 38.497825, acc = 0.845000\n",
      "epoch [1004] L = 38.497711, acc = 0.845000\n",
      "epoch [1005] L = 38.497598, acc = 0.845000\n",
      "epoch [1006] L = 38.497484, acc = 0.845000\n",
      "epoch [1007] L = 38.497370, acc = 0.845000\n",
      "epoch [1008] L = 38.497257, acc = 0.845000\n",
      "epoch [1009] L = 38.497144, acc = 0.845000\n",
      "epoch [1010] L = 38.497031, acc = 0.845000\n",
      "epoch [1011] L = 38.496918, acc = 0.845000\n",
      "epoch [1012] L = 38.496805, acc = 0.845000\n",
      "epoch [1013] L = 38.496692, acc = 0.845000\n",
      "epoch [1014] L = 38.496580, acc = 0.845000\n",
      "epoch [1015] L = 38.496467, acc = 0.845000\n",
      "epoch [1016] L = 38.496355, acc = 0.845000\n",
      "epoch [1017] L = 38.496243, acc = 0.845000\n",
      "epoch [1018] L = 38.496131, acc = 0.845000\n",
      "epoch [1019] L = 38.496019, acc = 0.845000\n",
      "epoch [1020] L = 38.495907, acc = 0.845000\n",
      "epoch [1021] L = 38.495795, acc = 0.845000\n",
      "epoch [1022] L = 38.495683, acc = 0.845000\n",
      "epoch [1023] L = 38.495572, acc = 0.845000\n",
      "epoch [1024] L = 38.495461, acc = 0.845000\n",
      "epoch [1025] L = 38.495349, acc = 0.845000\n",
      "epoch [1026] L = 38.495238, acc = 0.845000\n",
      "epoch [1027] L = 38.495127, acc = 0.845000\n",
      "epoch [1028] L = 38.495016, acc = 0.845000\n",
      "epoch [1029] L = 38.494906, acc = 0.845000\n",
      "epoch [1030] L = 38.494795, acc = 0.845000\n",
      "epoch [1031] L = 38.494685, acc = 0.845000\n",
      "epoch [1032] L = 38.494574, acc = 0.845000\n",
      "epoch [1033] L = 38.494464, acc = 0.845000\n",
      "epoch [1034] L = 38.494354, acc = 0.845000\n",
      "epoch [1035] L = 38.494244, acc = 0.845000\n",
      "epoch [1036] L = 38.494134, acc = 0.845000\n",
      "epoch [1037] L = 38.494024, acc = 0.845000\n",
      "epoch [1038] L = 38.493914, acc = 0.845000\n",
      "epoch [1039] L = 38.493805, acc = 0.845000\n",
      "epoch [1040] L = 38.493695, acc = 0.845000\n",
      "epoch [1041] L = 38.493586, acc = 0.845000\n",
      "epoch [1042] L = 38.493477, acc = 0.845000\n",
      "epoch [1043] L = 38.493368, acc = 0.845000\n",
      "epoch [1044] L = 38.493259, acc = 0.845000\n",
      "epoch [1045] L = 38.493150, acc = 0.845000\n",
      "epoch [1046] L = 38.493041, acc = 0.845000\n",
      "epoch [1047] L = 38.492933, acc = 0.845000\n",
      "epoch [1048] L = 38.492824, acc = 0.845000\n",
      "epoch [1049] L = 38.492716, acc = 0.845000\n",
      "epoch [1050] L = 38.492608, acc = 0.845000\n",
      "epoch [1051] L = 38.492499, acc = 0.845000\n",
      "epoch [1052] L = 38.492391, acc = 0.845000\n",
      "epoch [1053] L = 38.492283, acc = 0.845000\n",
      "epoch [1054] L = 38.492176, acc = 0.845000\n",
      "epoch [1055] L = 38.492068, acc = 0.845000\n",
      "epoch [1056] L = 38.491960, acc = 0.845000\n",
      "epoch [1057] L = 38.491853, acc = 0.845000\n",
      "epoch [1058] L = 38.491746, acc = 0.845000\n",
      "epoch [1059] L = 38.491638, acc = 0.845000\n",
      "epoch [1060] L = 38.491531, acc = 0.845000\n",
      "epoch [1061] L = 38.491424, acc = 0.845000\n",
      "epoch [1062] L = 38.491317, acc = 0.845000\n",
      "epoch [1063] L = 38.491211, acc = 0.845000\n",
      "epoch [1064] L = 38.491104, acc = 0.845000\n",
      "epoch [1065] L = 38.490997, acc = 0.845000\n",
      "epoch [1066] L = 38.490891, acc = 0.845000\n",
      "epoch [1067] L = 38.490785, acc = 0.845000\n",
      "epoch [1068] L = 38.490678, acc = 0.845000\n",
      "epoch [1069] L = 38.490572, acc = 0.845000\n",
      "epoch [1070] L = 38.490466, acc = 0.845000\n",
      "epoch [1071] L = 38.490360, acc = 0.845000\n",
      "epoch [1072] L = 38.490255, acc = 0.845000\n",
      "epoch [1073] L = 38.490149, acc = 0.845000\n",
      "epoch [1074] L = 38.490043, acc = 0.845000\n",
      "epoch [1075] L = 38.489938, acc = 0.845000\n",
      "epoch [1076] L = 38.489833, acc = 0.845000\n",
      "epoch [1077] L = 38.489727, acc = 0.845000\n",
      "epoch [1078] L = 38.489622, acc = 0.845000\n",
      "epoch [1079] L = 38.489517, acc = 0.845000\n",
      "epoch [1080] L = 38.489412, acc = 0.845000\n",
      "epoch [1081] L = 38.489307, acc = 0.845000\n",
      "epoch [1082] L = 38.489203, acc = 0.845000\n",
      "epoch [1083] L = 38.489098, acc = 0.845000\n",
      "epoch [1084] L = 38.488994, acc = 0.845000\n",
      "epoch [1085] L = 38.488889, acc = 0.845000\n",
      "epoch [1086] L = 38.488785, acc = 0.845000\n",
      "epoch [1087] L = 38.488681, acc = 0.845000\n",
      "epoch [1088] L = 38.488577, acc = 0.845000\n",
      "epoch [1089] L = 38.488473, acc = 0.845000\n",
      "epoch [1090] L = 38.488369, acc = 0.845000\n",
      "epoch [1091] L = 38.488265, acc = 0.845000\n",
      "epoch [1092] L = 38.488162, acc = 0.845000\n",
      "epoch [1093] L = 38.488058, acc = 0.845000\n",
      "epoch [1094] L = 38.487955, acc = 0.845000\n",
      "epoch [1095] L = 38.487852, acc = 0.845000\n",
      "epoch [1096] L = 38.487748, acc = 0.845000\n",
      "epoch [1097] L = 38.487645, acc = 0.845000\n",
      "epoch [1098] L = 38.487542, acc = 0.845000\n",
      "epoch [1099] L = 38.487439, acc = 0.845000\n",
      "epoch [1100] L = 38.487337, acc = 0.845000\n",
      "epoch [1101] L = 38.487234, acc = 0.845000\n",
      "epoch [1102] L = 38.487131, acc = 0.845000\n",
      "epoch [1103] L = 38.487029, acc = 0.845000\n",
      "epoch [1104] L = 38.486927, acc = 0.845000\n",
      "epoch [1105] L = 38.486824, acc = 0.845000\n",
      "epoch [1106] L = 38.486722, acc = 0.845000\n",
      "epoch [1107] L = 38.486620, acc = 0.845000\n",
      "epoch [1108] L = 38.486518, acc = 0.845000\n",
      "epoch [1109] L = 38.486416, acc = 0.845000\n",
      "epoch [1110] L = 38.486314, acc = 0.845000\n",
      "epoch [1111] L = 38.486213, acc = 0.845000\n",
      "epoch [1112] L = 38.486111, acc = 0.845000\n",
      "epoch [1113] L = 38.486010, acc = 0.845000\n",
      "epoch [1114] L = 38.485908, acc = 0.845000\n",
      "epoch [1115] L = 38.485807, acc = 0.845000\n",
      "epoch [1116] L = 38.485706, acc = 0.845000\n",
      "epoch [1117] L = 38.485605, acc = 0.845000\n",
      "epoch [1118] L = 38.485504, acc = 0.845000\n",
      "epoch [1119] L = 38.485403, acc = 0.845000\n",
      "epoch [1120] L = 38.485302, acc = 0.845000\n",
      "epoch [1121] L = 38.485202, acc = 0.845000\n",
      "epoch [1122] L = 38.485101, acc = 0.845000\n",
      "epoch [1123] L = 38.485001, acc = 0.845000\n",
      "epoch [1124] L = 38.484900, acc = 0.845000\n",
      "epoch [1125] L = 38.484800, acc = 0.845000\n",
      "epoch [1126] L = 38.484700, acc = 0.845000\n",
      "epoch [1127] L = 38.484600, acc = 0.845000\n",
      "epoch [1128] L = 38.484500, acc = 0.845000\n",
      "epoch [1129] L = 38.484400, acc = 0.845000\n",
      "epoch [1130] L = 38.484300, acc = 0.845000\n",
      "epoch [1131] L = 38.484201, acc = 0.845000\n",
      "epoch [1132] L = 38.484101, acc = 0.845000\n",
      "epoch [1133] L = 38.484002, acc = 0.845000\n",
      "epoch [1134] L = 38.483902, acc = 0.845000\n",
      "epoch [1135] L = 38.483803, acc = 0.845000\n",
      "epoch [1136] L = 38.483704, acc = 0.845000\n",
      "epoch [1137] L = 38.483605, acc = 0.845000\n",
      "epoch [1138] L = 38.483506, acc = 0.845000\n",
      "epoch [1139] L = 38.483407, acc = 0.845000\n",
      "epoch [1140] L = 38.483308, acc = 0.845000\n",
      "epoch [1141] L = 38.483210, acc = 0.845000\n",
      "epoch [1142] L = 38.483111, acc = 0.845000\n",
      "epoch [1143] L = 38.483012, acc = 0.845000\n",
      "epoch [1144] L = 38.482914, acc = 0.845000\n",
      "epoch [1145] L = 38.482816, acc = 0.845000\n",
      "epoch [1146] L = 38.482717, acc = 0.845000\n",
      "epoch [1147] L = 38.482619, acc = 0.845000\n",
      "epoch [1148] L = 38.482521, acc = 0.845000\n",
      "epoch [1149] L = 38.482423, acc = 0.845000\n",
      "epoch [1150] L = 38.482326, acc = 0.845000\n",
      "epoch [1151] L = 38.482228, acc = 0.845000\n",
      "epoch [1152] L = 38.482130, acc = 0.845000\n",
      "epoch [1153] L = 38.482033, acc = 0.845000\n",
      "epoch [1154] L = 38.481935, acc = 0.845000\n",
      "epoch [1155] L = 38.481838, acc = 0.845000\n",
      "epoch [1156] L = 38.481740, acc = 0.845000\n",
      "epoch [1157] L = 38.481643, acc = 0.845000\n",
      "epoch [1158] L = 38.481546, acc = 0.845000\n",
      "epoch [1159] L = 38.481449, acc = 0.845000\n",
      "epoch [1160] L = 38.481352, acc = 0.845000\n",
      "epoch [1161] L = 38.481255, acc = 0.845000\n",
      "epoch [1162] L = 38.481159, acc = 0.845000\n",
      "epoch [1163] L = 38.481062, acc = 0.845000\n",
      "epoch [1164] L = 38.480965, acc = 0.845000\n",
      "epoch [1165] L = 38.480869, acc = 0.845000\n",
      "epoch [1166] L = 38.480773, acc = 0.845000\n",
      "epoch [1167] L = 38.480676, acc = 0.845000\n",
      "epoch [1168] L = 38.480580, acc = 0.845000\n",
      "epoch [1169] L = 38.480484, acc = 0.845000\n",
      "epoch [1170] L = 38.480388, acc = 0.845000\n",
      "epoch [1171] L = 38.480292, acc = 0.845000\n",
      "epoch [1172] L = 38.480196, acc = 0.845000\n",
      "epoch [1173] L = 38.480101, acc = 0.845000\n",
      "epoch [1174] L = 38.480005, acc = 0.845000\n",
      "epoch [1175] L = 38.479909, acc = 0.845000\n",
      "epoch [1176] L = 38.479814, acc = 0.845000\n",
      "epoch [1177] L = 38.479719, acc = 0.845000\n",
      "epoch [1178] L = 38.479623, acc = 0.845000\n",
      "epoch [1179] L = 38.479528, acc = 0.845000\n",
      "epoch [1180] L = 38.479433, acc = 0.845000\n",
      "epoch [1181] L = 38.479338, acc = 0.845000\n",
      "epoch [1182] L = 38.479243, acc = 0.845000\n",
      "epoch [1183] L = 38.479148, acc = 0.845000\n",
      "epoch [1184] L = 38.479053, acc = 0.845000\n",
      "epoch [1185] L = 38.478959, acc = 0.845000\n",
      "epoch [1186] L = 38.478864, acc = 0.845000\n",
      "epoch [1187] L = 38.478770, acc = 0.845000\n",
      "epoch [1188] L = 38.478675, acc = 0.845000\n",
      "epoch [1189] L = 38.478581, acc = 0.845000\n",
      "epoch [1190] L = 38.478487, acc = 0.845000\n",
      "epoch [1191] L = 38.478393, acc = 0.845000\n",
      "epoch [1192] L = 38.478298, acc = 0.845000\n",
      "epoch [1193] L = 38.478205, acc = 0.845000\n",
      "epoch [1194] L = 38.478111, acc = 0.845000\n",
      "epoch [1195] L = 38.478017, acc = 0.845000\n",
      "epoch [1196] L = 38.477923, acc = 0.845000\n",
      "epoch [1197] L = 38.477829, acc = 0.845000\n",
      "epoch [1198] L = 38.477736, acc = 0.845000\n",
      "epoch [1199] L = 38.477642, acc = 0.845000\n",
      "epoch [1200] L = 38.477549, acc = 0.845000\n",
      "epoch [1201] L = 38.477456, acc = 0.845000\n",
      "epoch [1202] L = 38.477363, acc = 0.845000\n",
      "epoch [1203] L = 38.477269, acc = 0.845000\n",
      "epoch [1204] L = 38.477176, acc = 0.845000\n",
      "epoch [1205] L = 38.477083, acc = 0.845000\n",
      "epoch [1206] L = 38.476991, acc = 0.845000\n",
      "epoch [1207] L = 38.476898, acc = 0.845000\n",
      "epoch [1208] L = 38.476805, acc = 0.845000\n",
      "epoch [1209] L = 38.476712, acc = 0.845000\n",
      "epoch [1210] L = 38.476620, acc = 0.845000\n",
      "epoch [1211] L = 38.476528, acc = 0.845000\n",
      "epoch [1212] L = 38.476435, acc = 0.845000\n",
      "epoch [1213] L = 38.476343, acc = 0.845000\n",
      "epoch [1214] L = 38.476251, acc = 0.845000\n",
      "epoch [1215] L = 38.476159, acc = 0.845000\n",
      "epoch [1216] L = 38.476067, acc = 0.845000\n",
      "epoch [1217] L = 38.475975, acc = 0.845000\n",
      "epoch [1218] L = 38.475883, acc = 0.845000\n",
      "epoch [1219] L = 38.475791, acc = 0.845000\n",
      "epoch [1220] L = 38.475699, acc = 0.845000\n",
      "epoch [1221] L = 38.475608, acc = 0.845000\n",
      "epoch [1222] L = 38.475516, acc = 0.845000\n",
      "epoch [1223] L = 38.475425, acc = 0.845000\n",
      "epoch [1224] L = 38.475333, acc = 0.845000\n",
      "epoch [1225] L = 38.475242, acc = 0.845000\n",
      "epoch [1226] L = 38.475151, acc = 0.845000\n",
      "epoch [1227] L = 38.475060, acc = 0.845000\n",
      "epoch [1228] L = 38.474968, acc = 0.845000\n",
      "epoch [1229] L = 38.474877, acc = 0.845000\n",
      "epoch [1230] L = 38.474787, acc = 0.845000\n",
      "epoch [1231] L = 38.474696, acc = 0.845000\n",
      "epoch [1232] L = 38.474605, acc = 0.845000\n",
      "epoch [1233] L = 38.474514, acc = 0.845000\n",
      "epoch [1234] L = 38.474424, acc = 0.845000\n",
      "epoch [1235] L = 38.474333, acc = 0.845000\n",
      "epoch [1236] L = 38.474243, acc = 0.845000\n",
      "epoch [1237] L = 38.474153, acc = 0.845000\n",
      "epoch [1238] L = 38.474062, acc = 0.845000\n",
      "epoch [1239] L = 38.473972, acc = 0.845000\n",
      "epoch [1240] L = 38.473882, acc = 0.845000\n",
      "epoch [1241] L = 38.473792, acc = 0.845000\n",
      "epoch [1242] L = 38.473702, acc = 0.845000\n",
      "epoch [1243] L = 38.473612, acc = 0.845000\n",
      "epoch [1244] L = 38.473522, acc = 0.845000\n",
      "epoch [1245] L = 38.473433, acc = 0.845000\n",
      "epoch [1246] L = 38.473343, acc = 0.845000\n",
      "epoch [1247] L = 38.473253, acc = 0.845000\n",
      "epoch [1248] L = 38.473164, acc = 0.845000\n",
      "epoch [1249] L = 38.473075, acc = 0.845000\n",
      "epoch [1250] L = 38.472985, acc = 0.845000\n",
      "epoch [1251] L = 38.472896, acc = 0.845000\n",
      "epoch [1252] L = 38.472807, acc = 0.845000\n",
      "epoch [1253] L = 38.472718, acc = 0.845000\n",
      "epoch [1254] L = 38.472629, acc = 0.845000\n",
      "epoch [1255] L = 38.472540, acc = 0.845000\n",
      "epoch [1256] L = 38.472451, acc = 0.845000\n",
      "epoch [1257] L = 38.472362, acc = 0.845000\n",
      "epoch [1258] L = 38.472273, acc = 0.845000\n",
      "epoch [1259] L = 38.472185, acc = 0.845000\n",
      "epoch [1260] L = 38.472096, acc = 0.845000\n",
      "epoch [1261] L = 38.472008, acc = 0.845000\n",
      "epoch [1262] L = 38.471919, acc = 0.845000\n",
      "epoch [1263] L = 38.471831, acc = 0.845000\n",
      "epoch [1264] L = 38.471743, acc = 0.845000\n",
      "epoch [1265] L = 38.471655, acc = 0.845000\n",
      "epoch [1266] L = 38.471567, acc = 0.845000\n",
      "epoch [1267] L = 38.471479, acc = 0.845000\n",
      "epoch [1268] L = 38.471391, acc = 0.845000\n",
      "epoch [1269] L = 38.471303, acc = 0.845000\n",
      "epoch [1270] L = 38.471215, acc = 0.845000\n",
      "epoch [1271] L = 38.471127, acc = 0.845000\n",
      "epoch [1272] L = 38.471040, acc = 0.845000\n",
      "epoch [1273] L = 38.470952, acc = 0.845000\n",
      "epoch [1274] L = 38.470864, acc = 0.845000\n",
      "epoch [1275] L = 38.470777, acc = 0.845000\n",
      "epoch [1276] L = 38.470690, acc = 0.845000\n",
      "epoch [1277] L = 38.470602, acc = 0.845000\n",
      "epoch [1278] L = 38.470515, acc = 0.845000\n",
      "epoch [1279] L = 38.470428, acc = 0.845000\n",
      "epoch [1280] L = 38.470341, acc = 0.845000\n",
      "epoch [1281] L = 38.470254, acc = 0.845000\n",
      "epoch [1282] L = 38.470167, acc = 0.845000\n",
      "epoch [1283] L = 38.470080, acc = 0.845000\n",
      "epoch [1284] L = 38.469993, acc = 0.845000\n",
      "epoch [1285] L = 38.469907, acc = 0.845000\n",
      "epoch [1286] L = 38.469820, acc = 0.845000\n",
      "epoch [1287] L = 38.469734, acc = 0.845000\n",
      "epoch [1288] L = 38.469647, acc = 0.845000\n",
      "epoch [1289] L = 38.469561, acc = 0.845000\n",
      "epoch [1290] L = 38.469474, acc = 0.845000\n",
      "epoch [1291] L = 38.469388, acc = 0.845000\n",
      "epoch [1292] L = 38.469302, acc = 0.845000\n",
      "epoch [1293] L = 38.469216, acc = 0.845000\n",
      "epoch [1294] L = 38.469130, acc = 0.845000\n",
      "epoch [1295] L = 38.469044, acc = 0.845000\n",
      "epoch [1296] L = 38.468958, acc = 0.845000\n",
      "epoch [1297] L = 38.468872, acc = 0.845000\n",
      "epoch [1298] L = 38.468786, acc = 0.845000\n",
      "epoch [1299] L = 38.468700, acc = 0.845000\n",
      "epoch [1300] L = 38.468615, acc = 0.845000\n",
      "epoch [1301] L = 38.468529, acc = 0.845000\n",
      "epoch [1302] L = 38.468444, acc = 0.845000\n",
      "epoch [1303] L = 38.468358, acc = 0.845000\n",
      "epoch [1304] L = 38.468273, acc = 0.845000\n",
      "epoch [1305] L = 38.468188, acc = 0.845000\n",
      "epoch [1306] L = 38.468103, acc = 0.845000\n",
      "epoch [1307] L = 38.468017, acc = 0.845000\n",
      "epoch [1308] L = 38.467932, acc = 0.845000\n",
      "epoch [1309] L = 38.467847, acc = 0.845000\n",
      "epoch [1310] L = 38.467762, acc = 0.845000\n",
      "epoch [1311] L = 38.467678, acc = 0.845000\n",
      "epoch [1312] L = 38.467593, acc = 0.845000\n",
      "epoch [1313] L = 38.467508, acc = 0.845000\n",
      "epoch [1314] L = 38.467423, acc = 0.845000\n",
      "epoch [1315] L = 38.467339, acc = 0.845000\n",
      "epoch [1316] L = 38.467254, acc = 0.845000\n",
      "epoch [1317] L = 38.467170, acc = 0.845000\n",
      "epoch [1318] L = 38.467086, acc = 0.845000\n",
      "epoch [1319] L = 38.467001, acc = 0.845000\n",
      "epoch [1320] L = 38.466917, acc = 0.845000\n",
      "epoch [1321] L = 38.466833, acc = 0.845000\n",
      "epoch [1322] L = 38.466749, acc = 0.845000\n",
      "epoch [1323] L = 38.466665, acc = 0.845000\n",
      "epoch [1324] L = 38.466581, acc = 0.845000\n",
      "epoch [1325] L = 38.466497, acc = 0.845000\n",
      "epoch [1326] L = 38.466413, acc = 0.845000\n",
      "epoch [1327] L = 38.466329, acc = 0.845000\n",
      "epoch [1328] L = 38.466245, acc = 0.845000\n",
      "epoch [1329] L = 38.466162, acc = 0.845000\n",
      "epoch [1330] L = 38.466078, acc = 0.845000\n",
      "epoch [1331] L = 38.465995, acc = 0.845000\n",
      "epoch [1332] L = 38.465911, acc = 0.845000\n",
      "epoch [1333] L = 38.465828, acc = 0.845000\n",
      "epoch [1334] L = 38.465745, acc = 0.845000\n",
      "epoch [1335] L = 38.465661, acc = 0.845000\n",
      "epoch [1336] L = 38.465578, acc = 0.845000\n",
      "epoch [1337] L = 38.465495, acc = 0.845000\n",
      "epoch [1338] L = 38.465412, acc = 0.845000\n",
      "epoch [1339] L = 38.465329, acc = 0.845000\n",
      "epoch [1340] L = 38.465246, acc = 0.845000\n",
      "epoch [1341] L = 38.465163, acc = 0.845000\n",
      "epoch [1342] L = 38.465081, acc = 0.845000\n",
      "epoch [1343] L = 38.464998, acc = 0.845000\n",
      "epoch [1344] L = 38.464915, acc = 0.845000\n",
      "epoch [1345] L = 38.464833, acc = 0.845000\n",
      "epoch [1346] L = 38.464750, acc = 0.845000\n",
      "epoch [1347] L = 38.464668, acc = 0.845000\n",
      "epoch [1348] L = 38.464585, acc = 0.845000\n",
      "epoch [1349] L = 38.464503, acc = 0.845000\n",
      "epoch [1350] L = 38.464421, acc = 0.845000\n",
      "epoch [1351] L = 38.464339, acc = 0.845000\n",
      "epoch [1352] L = 38.464257, acc = 0.845000\n",
      "epoch [1353] L = 38.464174, acc = 0.845000\n",
      "epoch [1354] L = 38.464092, acc = 0.845000\n",
      "epoch [1355] L = 38.464011, acc = 0.845000\n",
      "epoch [1356] L = 38.463929, acc = 0.845000\n",
      "epoch [1357] L = 38.463847, acc = 0.845000\n",
      "epoch [1358] L = 38.463765, acc = 0.845000\n",
      "epoch [1359] L = 38.463683, acc = 0.845000\n",
      "epoch [1360] L = 38.463602, acc = 0.845000\n",
      "epoch [1361] L = 38.463520, acc = 0.845000\n",
      "epoch [1362] L = 38.463439, acc = 0.845000\n",
      "epoch [1363] L = 38.463357, acc = 0.845000\n",
      "epoch [1364] L = 38.463276, acc = 0.845000\n",
      "epoch [1365] L = 38.463195, acc = 0.845000\n",
      "epoch [1366] L = 38.463113, acc = 0.845000\n",
      "epoch [1367] L = 38.463032, acc = 0.845000\n",
      "epoch [1368] L = 38.462951, acc = 0.845000\n",
      "epoch [1369] L = 38.462870, acc = 0.845000\n",
      "epoch [1370] L = 38.462789, acc = 0.845000\n",
      "epoch [1371] L = 38.462708, acc = 0.845000\n",
      "epoch [1372] L = 38.462627, acc = 0.845000\n",
      "epoch [1373] L = 38.462547, acc = 0.845000\n",
      "epoch [1374] L = 38.462466, acc = 0.845000\n",
      "epoch [1375] L = 38.462385, acc = 0.845000\n",
      "epoch [1376] L = 38.462305, acc = 0.845000\n",
      "epoch [1377] L = 38.462224, acc = 0.845000\n",
      "epoch [1378] L = 38.462144, acc = 0.845000\n",
      "epoch [1379] L = 38.462063, acc = 0.845000\n",
      "epoch [1380] L = 38.461983, acc = 0.845000\n",
      "epoch [1381] L = 38.461902, acc = 0.845000\n",
      "epoch [1382] L = 38.461822, acc = 0.845000\n",
      "epoch [1383] L = 38.461742, acc = 0.845000\n",
      "epoch [1384] L = 38.461662, acc = 0.845000\n",
      "epoch [1385] L = 38.461582, acc = 0.845000\n",
      "epoch [1386] L = 38.461502, acc = 0.845000\n",
      "epoch [1387] L = 38.461422, acc = 0.845000\n",
      "epoch [1388] L = 38.461342, acc = 0.845000\n",
      "epoch [1389] L = 38.461262, acc = 0.845000\n",
      "epoch [1390] L = 38.461182, acc = 0.845000\n",
      "epoch [1391] L = 38.461103, acc = 0.845000\n",
      "epoch [1392] L = 38.461023, acc = 0.845000\n",
      "epoch [1393] L = 38.460944, acc = 0.845000\n",
      "epoch [1394] L = 38.460864, acc = 0.845000\n",
      "epoch [1395] L = 38.460785, acc = 0.845000\n",
      "epoch [1396] L = 38.460705, acc = 0.845000\n",
      "epoch [1397] L = 38.460626, acc = 0.845000\n",
      "epoch [1398] L = 38.460547, acc = 0.845000\n",
      "epoch [1399] L = 38.460467, acc = 0.845000\n",
      "epoch [1400] L = 38.460388, acc = 0.845000\n",
      "epoch [1401] L = 38.460309, acc = 0.845000\n",
      "epoch [1402] L = 38.460230, acc = 0.845000\n",
      "epoch [1403] L = 38.460151, acc = 0.845000\n",
      "epoch [1404] L = 38.460072, acc = 0.845000\n",
      "epoch [1405] L = 38.459993, acc = 0.845000\n",
      "epoch [1406] L = 38.459914, acc = 0.845000\n",
      "epoch [1407] L = 38.459836, acc = 0.845000\n",
      "epoch [1408] L = 38.459757, acc = 0.845000\n",
      "epoch [1409] L = 38.459678, acc = 0.845000\n",
      "epoch [1410] L = 38.459600, acc = 0.845000\n",
      "epoch [1411] L = 38.459521, acc = 0.845000\n",
      "epoch [1412] L = 38.459443, acc = 0.845000\n",
      "epoch [1413] L = 38.459364, acc = 0.845000\n",
      "epoch [1414] L = 38.459286, acc = 0.845000\n",
      "epoch [1415] L = 38.459208, acc = 0.845000\n",
      "epoch [1416] L = 38.459130, acc = 0.845000\n",
      "epoch [1417] L = 38.459051, acc = 0.845000\n",
      "epoch [1418] L = 38.458973, acc = 0.845000\n",
      "epoch [1419] L = 38.458895, acc = 0.845000\n",
      "epoch [1420] L = 38.458817, acc = 0.845000\n",
      "epoch [1421] L = 38.458739, acc = 0.845000\n",
      "epoch [1422] L = 38.458661, acc = 0.845000\n",
      "epoch [1423] L = 38.458584, acc = 0.845000\n",
      "epoch [1424] L = 38.458506, acc = 0.845000\n",
      "epoch [1425] L = 38.458428, acc = 0.845000\n",
      "epoch [1426] L = 38.458350, acc = 0.845000\n",
      "epoch [1427] L = 38.458273, acc = 0.845000\n",
      "epoch [1428] L = 38.458195, acc = 0.845000\n",
      "epoch [1429] L = 38.458118, acc = 0.845000\n",
      "epoch [1430] L = 38.458040, acc = 0.845000\n",
      "epoch [1431] L = 38.457963, acc = 0.845000\n",
      "epoch [1432] L = 38.457886, acc = 0.845000\n",
      "epoch [1433] L = 38.457808, acc = 0.845000\n",
      "epoch [1434] L = 38.457731, acc = 0.845000\n",
      "epoch [1435] L = 38.457654, acc = 0.845000\n",
      "epoch [1436] L = 38.457577, acc = 0.845000\n",
      "epoch [1437] L = 38.457500, acc = 0.845000\n",
      "epoch [1438] L = 38.457423, acc = 0.845000\n",
      "epoch [1439] L = 38.457346, acc = 0.845000\n",
      "epoch [1440] L = 38.457269, acc = 0.845000\n",
      "epoch [1441] L = 38.457192, acc = 0.845000\n",
      "epoch [1442] L = 38.457116, acc = 0.845000\n",
      "epoch [1443] L = 38.457039, acc = 0.845000\n",
      "epoch [1444] L = 38.456962, acc = 0.845000\n",
      "epoch [1445] L = 38.456886, acc = 0.845000\n",
      "epoch [1446] L = 38.456809, acc = 0.845000\n",
      "epoch [1447] L = 38.456733, acc = 0.845000\n",
      "epoch [1448] L = 38.456656, acc = 0.845000\n",
      "epoch [1449] L = 38.456580, acc = 0.845000\n",
      "epoch [1450] L = 38.456504, acc = 0.845000\n",
      "epoch [1451] L = 38.456427, acc = 0.845000\n",
      "epoch [1452] L = 38.456351, acc = 0.845000\n",
      "epoch [1453] L = 38.456275, acc = 0.845000\n",
      "epoch [1454] L = 38.456199, acc = 0.845000\n",
      "epoch [1455] L = 38.456123, acc = 0.845000\n",
      "epoch [1456] L = 38.456047, acc = 0.845000\n",
      "epoch [1457] L = 38.455971, acc = 0.845000\n",
      "epoch [1458] L = 38.455895, acc = 0.845000\n",
      "epoch [1459] L = 38.455819, acc = 0.845000\n",
      "epoch [1460] L = 38.455743, acc = 0.845000\n",
      "epoch [1461] L = 38.455668, acc = 0.845000\n",
      "epoch [1462] L = 38.455592, acc = 0.845000\n",
      "epoch [1463] L = 38.455516, acc = 0.845000\n",
      "epoch [1464] L = 38.455441, acc = 0.845000\n",
      "epoch [1465] L = 38.455365, acc = 0.845000\n",
      "epoch [1466] L = 38.455290, acc = 0.845000\n",
      "epoch [1467] L = 38.455214, acc = 0.845000\n",
      "epoch [1468] L = 38.455139, acc = 0.845000\n",
      "epoch [1469] L = 38.455064, acc = 0.845000\n",
      "epoch [1470] L = 38.454988, acc = 0.845000\n",
      "epoch [1471] L = 38.454913, acc = 0.845000\n",
      "epoch [1472] L = 38.454838, acc = 0.845000\n",
      "epoch [1473] L = 38.454763, acc = 0.845000\n",
      "epoch [1474] L = 38.454688, acc = 0.845000\n",
      "epoch [1475] L = 38.454613, acc = 0.845000\n",
      "epoch [1476] L = 38.454538, acc = 0.845000\n",
      "epoch [1477] L = 38.454463, acc = 0.845000\n",
      "epoch [1478] L = 38.454388, acc = 0.845000\n",
      "epoch [1479] L = 38.454314, acc = 0.845000\n",
      "epoch [1480] L = 38.454239, acc = 0.845000\n",
      "epoch [1481] L = 38.454164, acc = 0.845000\n",
      "epoch [1482] L = 38.454090, acc = 0.845000\n",
      "epoch [1483] L = 38.454015, acc = 0.845000\n",
      "epoch [1484] L = 38.453941, acc = 0.845000\n",
      "epoch [1485] L = 38.453866, acc = 0.845000\n",
      "epoch [1486] L = 38.453792, acc = 0.845000\n",
      "epoch [1487] L = 38.453717, acc = 0.845000\n",
      "epoch [1488] L = 38.453643, acc = 0.845000\n",
      "epoch [1489] L = 38.453569, acc = 0.845000\n",
      "epoch [1490] L = 38.453494, acc = 0.845000\n",
      "epoch [1491] L = 38.453420, acc = 0.845000\n",
      "epoch [1492] L = 38.453346, acc = 0.845000\n",
      "epoch [1493] L = 38.453272, acc = 0.845000\n",
      "epoch [1494] L = 38.453198, acc = 0.845000\n",
      "epoch [1495] L = 38.453124, acc = 0.845000\n",
      "epoch [1496] L = 38.453050, acc = 0.845000\n",
      "epoch [1497] L = 38.452976, acc = 0.845000\n",
      "epoch [1498] L = 38.452903, acc = 0.845000\n",
      "epoch [1499] L = 38.452829, acc = 0.845000\n",
      "epoch [1500] L = 38.452755, acc = 0.845000\n",
      "epoch [1501] L = 38.452682, acc = 0.845000\n",
      "epoch [1502] L = 38.452608, acc = 0.845000\n",
      "epoch [1503] L = 38.452534, acc = 0.845000\n",
      "epoch [1504] L = 38.452461, acc = 0.845000\n",
      "epoch [1505] L = 38.452387, acc = 0.845000\n",
      "epoch [1506] L = 38.452314, acc = 0.845000\n",
      "epoch [1507] L = 38.452241, acc = 0.845000\n",
      "epoch [1508] L = 38.452167, acc = 0.845000\n",
      "epoch [1509] L = 38.452094, acc = 0.845000\n",
      "epoch [1510] L = 38.452021, acc = 0.845000\n",
      "epoch [1511] L = 38.451948, acc = 0.845000\n",
      "epoch [1512] L = 38.451875, acc = 0.845000\n",
      "epoch [1513] L = 38.451802, acc = 0.845000\n",
      "epoch [1514] L = 38.451729, acc = 0.845000\n",
      "epoch [1515] L = 38.451656, acc = 0.845000\n",
      "epoch [1516] L = 38.451583, acc = 0.845000\n",
      "epoch [1517] L = 38.451510, acc = 0.845000\n",
      "epoch [1518] L = 38.451437, acc = 0.845000\n",
      "epoch [1519] L = 38.451364, acc = 0.845000\n",
      "epoch [1520] L = 38.451292, acc = 0.845000\n",
      "epoch [1521] L = 38.451219, acc = 0.845000\n",
      "epoch [1522] L = 38.451146, acc = 0.845000\n",
      "epoch [1523] L = 38.451074, acc = 0.845000\n",
      "epoch [1524] L = 38.451001, acc = 0.845000\n",
      "epoch [1525] L = 38.450929, acc = 0.845000\n",
      "epoch [1526] L = 38.450856, acc = 0.845000\n",
      "epoch [1527] L = 38.450784, acc = 0.845000\n",
      "epoch [1528] L = 38.450712, acc = 0.845000\n",
      "epoch [1529] L = 38.450639, acc = 0.845000\n",
      "epoch [1530] L = 38.450567, acc = 0.845000\n",
      "epoch [1531] L = 38.450495, acc = 0.845000\n",
      "epoch [1532] L = 38.450423, acc = 0.845000\n",
      "epoch [1533] L = 38.450351, acc = 0.845000\n",
      "epoch [1534] L = 38.450279, acc = 0.845000\n",
      "epoch [1535] L = 38.450207, acc = 0.845000\n",
      "epoch [1536] L = 38.450135, acc = 0.845000\n",
      "epoch [1537] L = 38.450063, acc = 0.845000\n",
      "epoch [1538] L = 38.449991, acc = 0.845000\n",
      "epoch [1539] L = 38.449919, acc = 0.845000\n",
      "epoch [1540] L = 38.449847, acc = 0.845000\n",
      "epoch [1541] L = 38.449776, acc = 0.845000\n",
      "epoch [1542] L = 38.449704, acc = 0.845000\n",
      "epoch [1543] L = 38.449633, acc = 0.845000\n",
      "epoch [1544] L = 38.449561, acc = 0.845000\n",
      "epoch [1545] L = 38.449489, acc = 0.845000\n",
      "epoch [1546] L = 38.449418, acc = 0.845000\n",
      "epoch [1547] L = 38.449347, acc = 0.845000\n",
      "epoch [1548] L = 38.449275, acc = 0.845000\n",
      "epoch [1549] L = 38.449204, acc = 0.845000\n",
      "epoch [1550] L = 38.449133, acc = 0.845000\n",
      "epoch [1551] L = 38.449061, acc = 0.845000\n",
      "epoch [1552] L = 38.448990, acc = 0.845000\n",
      "epoch [1553] L = 38.448919, acc = 0.845000\n",
      "epoch [1554] L = 38.448848, acc = 0.845000\n",
      "epoch [1555] L = 38.448777, acc = 0.845000\n",
      "epoch [1556] L = 38.448706, acc = 0.845000\n",
      "epoch [1557] L = 38.448635, acc = 0.845000\n",
      "epoch [1558] L = 38.448564, acc = 0.845000\n",
      "epoch [1559] L = 38.448493, acc = 0.845000\n",
      "epoch [1560] L = 38.448422, acc = 0.845000\n",
      "epoch [1561] L = 38.448351, acc = 0.845000\n",
      "epoch [1562] L = 38.448281, acc = 0.845000\n",
      "epoch [1563] L = 38.448210, acc = 0.845000\n",
      "epoch [1564] L = 38.448139, acc = 0.845000\n",
      "epoch [1565] L = 38.448069, acc = 0.845000\n",
      "epoch [1566] L = 38.447998, acc = 0.845000\n",
      "epoch [1567] L = 38.447928, acc = 0.845000\n",
      "epoch [1568] L = 38.447857, acc = 0.845000\n",
      "epoch [1569] L = 38.447787, acc = 0.845000\n",
      "epoch [1570] L = 38.447717, acc = 0.845000\n",
      "epoch [1571] L = 38.447646, acc = 0.845000\n",
      "epoch [1572] L = 38.447576, acc = 0.845000\n",
      "epoch [1573] L = 38.447506, acc = 0.845000\n",
      "epoch [1574] L = 38.447436, acc = 0.845000\n",
      "epoch [1575] L = 38.447365, acc = 0.845000\n",
      "epoch [1576] L = 38.447295, acc = 0.845000\n",
      "epoch [1577] L = 38.447225, acc = 0.845000\n",
      "epoch [1578] L = 38.447155, acc = 0.845000\n",
      "epoch [1579] L = 38.447085, acc = 0.845000\n",
      "epoch [1580] L = 38.447015, acc = 0.845000\n",
      "epoch [1581] L = 38.446945, acc = 0.845000\n",
      "epoch [1582] L = 38.446876, acc = 0.845000\n",
      "epoch [1583] L = 38.446806, acc = 0.845000\n",
      "epoch [1584] L = 38.446736, acc = 0.845000\n",
      "epoch [1585] L = 38.446666, acc = 0.845000\n",
      "epoch [1586] L = 38.446597, acc = 0.845000\n",
      "epoch [1587] L = 38.446527, acc = 0.845000\n",
      "epoch [1588] L = 38.446458, acc = 0.845000\n",
      "epoch [1589] L = 38.446388, acc = 0.845000\n",
      "epoch [1590] L = 38.446319, acc = 0.845000\n",
      "epoch [1591] L = 38.446249, acc = 0.845000\n",
      "epoch [1592] L = 38.446180, acc = 0.845000\n",
      "epoch [1593] L = 38.446110, acc = 0.845000\n",
      "epoch [1594] L = 38.446041, acc = 0.845000\n",
      "epoch [1595] L = 38.445972, acc = 0.845000\n",
      "epoch [1596] L = 38.445903, acc = 0.845000\n",
      "epoch [1597] L = 38.445833, acc = 0.845000\n",
      "epoch [1598] L = 38.445764, acc = 0.845000\n",
      "epoch [1599] L = 38.445695, acc = 0.845000\n",
      "epoch [1600] L = 38.445626, acc = 0.845000\n",
      "epoch [1601] L = 38.445557, acc = 0.845000\n",
      "epoch [1602] L = 38.445488, acc = 0.845000\n",
      "epoch [1603] L = 38.445419, acc = 0.845000\n",
      "epoch [1604] L = 38.445350, acc = 0.845000\n",
      "epoch [1605] L = 38.445282, acc = 0.845000\n",
      "epoch [1606] L = 38.445213, acc = 0.845000\n",
      "epoch [1607] L = 38.445144, acc = 0.845000\n",
      "epoch [1608] L = 38.445075, acc = 0.845000\n",
      "epoch [1609] L = 38.445007, acc = 0.845000\n",
      "epoch [1610] L = 38.444938, acc = 0.845000\n",
      "epoch [1611] L = 38.444870, acc = 0.845000\n",
      "epoch [1612] L = 38.444801, acc = 0.845000\n",
      "epoch [1613] L = 38.444733, acc = 0.845000\n",
      "epoch [1614] L = 38.444664, acc = 0.845000\n",
      "epoch [1615] L = 38.444596, acc = 0.845000\n",
      "epoch [1616] L = 38.444527, acc = 0.845000\n",
      "epoch [1617] L = 38.444459, acc = 0.845000\n",
      "epoch [1618] L = 38.444391, acc = 0.845000\n",
      "epoch [1619] L = 38.444323, acc = 0.845000\n",
      "epoch [1620] L = 38.444254, acc = 0.845000\n",
      "epoch [1621] L = 38.444186, acc = 0.845000\n",
      "epoch [1622] L = 38.444118, acc = 0.845000\n",
      "epoch [1623] L = 38.444050, acc = 0.845000\n",
      "epoch [1624] L = 38.443982, acc = 0.845000\n",
      "epoch [1625] L = 38.443914, acc = 0.845000\n",
      "epoch [1626] L = 38.443846, acc = 0.845000\n",
      "epoch [1627] L = 38.443778, acc = 0.845000\n",
      "epoch [1628] L = 38.443710, acc = 0.845000\n",
      "epoch [1629] L = 38.443642, acc = 0.845000\n",
      "epoch [1630] L = 38.443575, acc = 0.845000\n",
      "epoch [1631] L = 38.443507, acc = 0.845000\n",
      "epoch [1632] L = 38.443439, acc = 0.845000\n",
      "epoch [1633] L = 38.443372, acc = 0.845000\n",
      "epoch [1634] L = 38.443304, acc = 0.845000\n",
      "epoch [1635] L = 38.443236, acc = 0.845000\n",
      "epoch [1636] L = 38.443169, acc = 0.845000\n",
      "epoch [1637] L = 38.443101, acc = 0.845000\n",
      "epoch [1638] L = 38.443034, acc = 0.845000\n",
      "epoch [1639] L = 38.442967, acc = 0.845000\n",
      "epoch [1640] L = 38.442899, acc = 0.845000\n",
      "epoch [1641] L = 38.442832, acc = 0.845000\n",
      "epoch [1642] L = 38.442765, acc = 0.845000\n",
      "epoch [1643] L = 38.442697, acc = 0.845000\n",
      "epoch [1644] L = 38.442630, acc = 0.845000\n",
      "epoch [1645] L = 38.442563, acc = 0.845000\n",
      "epoch [1646] L = 38.442496, acc = 0.845000\n",
      "epoch [1647] L = 38.442429, acc = 0.845000\n",
      "epoch [1648] L = 38.442362, acc = 0.845000\n",
      "epoch [1649] L = 38.442295, acc = 0.845000\n",
      "epoch [1650] L = 38.442228, acc = 0.845000\n",
      "epoch [1651] L = 38.442161, acc = 0.845000\n",
      "epoch [1652] L = 38.442094, acc = 0.845000\n",
      "epoch [1653] L = 38.442027, acc = 0.845000\n",
      "epoch [1654] L = 38.441960, acc = 0.845000\n",
      "epoch [1655] L = 38.441894, acc = 0.845000\n",
      "epoch [1656] L = 38.441827, acc = 0.845000\n",
      "epoch [1657] L = 38.441760, acc = 0.845000\n",
      "epoch [1658] L = 38.441693, acc = 0.845000\n",
      "epoch [1659] L = 38.441627, acc = 0.845000\n",
      "epoch [1660] L = 38.441560, acc = 0.845000\n",
      "epoch [1661] L = 38.441494, acc = 0.845000\n",
      "epoch [1662] L = 38.441427, acc = 0.845000\n",
      "epoch [1663] L = 38.441361, acc = 0.845000\n",
      "epoch [1664] L = 38.441294, acc = 0.845000\n",
      "epoch [1665] L = 38.441228, acc = 0.845000\n",
      "epoch [1666] L = 38.441162, acc = 0.845000\n",
      "epoch [1667] L = 38.441095, acc = 0.845000\n",
      "epoch [1668] L = 38.441029, acc = 0.845000\n",
      "epoch [1669] L = 38.440963, acc = 0.845000\n",
      "epoch [1670] L = 38.440897, acc = 0.845000\n",
      "epoch [1671] L = 38.440831, acc = 0.845000\n",
      "epoch [1672] L = 38.440765, acc = 0.845000\n",
      "epoch [1673] L = 38.440699, acc = 0.845000\n",
      "epoch [1674] L = 38.440633, acc = 0.845000\n",
      "epoch [1675] L = 38.440567, acc = 0.845000\n",
      "epoch [1676] L = 38.440501, acc = 0.845000\n",
      "epoch [1677] L = 38.440435, acc = 0.845000\n",
      "epoch [1678] L = 38.440369, acc = 0.845000\n",
      "epoch [1679] L = 38.440303, acc = 0.845000\n",
      "epoch [1680] L = 38.440237, acc = 0.845000\n",
      "epoch [1681] L = 38.440172, acc = 0.845000\n",
      "epoch [1682] L = 38.440106, acc = 0.845000\n",
      "epoch [1683] L = 38.440040, acc = 0.845000\n",
      "epoch [1684] L = 38.439975, acc = 0.845000\n",
      "epoch [1685] L = 38.439909, acc = 0.845000\n",
      "epoch [1686] L = 38.439843, acc = 0.845000\n",
      "epoch [1687] L = 38.439778, acc = 0.845000\n",
      "epoch [1688] L = 38.439712, acc = 0.845000\n",
      "epoch [1689] L = 38.439647, acc = 0.845000\n",
      "epoch [1690] L = 38.439582, acc = 0.845000\n",
      "epoch [1691] L = 38.439516, acc = 0.845000\n",
      "epoch [1692] L = 38.439451, acc = 0.845000\n",
      "epoch [1693] L = 38.439386, acc = 0.845000\n",
      "epoch [1694] L = 38.439320, acc = 0.845000\n",
      "epoch [1695] L = 38.439255, acc = 0.845000\n",
      "epoch [1696] L = 38.439190, acc = 0.845000\n",
      "epoch [1697] L = 38.439125, acc = 0.845000\n",
      "epoch [1698] L = 38.439060, acc = 0.845000\n",
      "epoch [1699] L = 38.438995, acc = 0.845000\n",
      "epoch [1700] L = 38.438930, acc = 0.845000\n",
      "epoch [1701] L = 38.438865, acc = 0.845000\n",
      "epoch [1702] L = 38.438800, acc = 0.845000\n",
      "epoch [1703] L = 38.438735, acc = 0.845000\n",
      "epoch [1704] L = 38.438670, acc = 0.845000\n",
      "epoch [1705] L = 38.438605, acc = 0.845000\n",
      "epoch [1706] L = 38.438540, acc = 0.845000\n",
      "epoch [1707] L = 38.438476, acc = 0.845000\n",
      "epoch [1708] L = 38.438411, acc = 0.845000\n",
      "epoch [1709] L = 38.438346, acc = 0.845000\n",
      "epoch [1710] L = 38.438281, acc = 0.845000\n",
      "epoch [1711] L = 38.438217, acc = 0.845000\n",
      "epoch [1712] L = 38.438152, acc = 0.845000\n",
      "epoch [1713] L = 38.438088, acc = 0.845000\n",
      "epoch [1714] L = 38.438023, acc = 0.845000\n",
      "epoch [1715] L = 38.437959, acc = 0.845000\n",
      "epoch [1716] L = 38.437894, acc = 0.845000\n",
      "epoch [1717] L = 38.437830, acc = 0.845000\n",
      "epoch [1718] L = 38.437766, acc = 0.845000\n",
      "epoch [1719] L = 38.437701, acc = 0.845000\n",
      "epoch [1720] L = 38.437637, acc = 0.845000\n",
      "epoch [1721] L = 38.437573, acc = 0.845000\n",
      "epoch [1722] L = 38.437508, acc = 0.845000\n",
      "epoch [1723] L = 38.437444, acc = 0.845000\n",
      "epoch [1724] L = 38.437380, acc = 0.845000\n",
      "epoch [1725] L = 38.437316, acc = 0.845000\n",
      "epoch [1726] L = 38.437252, acc = 0.845000\n",
      "epoch [1727] L = 38.437188, acc = 0.845000\n",
      "epoch [1728] L = 38.437124, acc = 0.845000\n",
      "epoch [1729] L = 38.437060, acc = 0.845000\n",
      "epoch [1730] L = 38.436996, acc = 0.845000\n",
      "epoch [1731] L = 38.436932, acc = 0.845000\n",
      "epoch [1732] L = 38.436868, acc = 0.845000\n",
      "epoch [1733] L = 38.436805, acc = 0.845000\n",
      "epoch [1734] L = 38.436741, acc = 0.845000\n",
      "epoch [1735] L = 38.436677, acc = 0.845000\n",
      "epoch [1736] L = 38.436613, acc = 0.845000\n",
      "epoch [1737] L = 38.436550, acc = 0.845000\n",
      "epoch [1738] L = 38.436486, acc = 0.845000\n",
      "epoch [1739] L = 38.436422, acc = 0.845000\n",
      "epoch [1740] L = 38.436359, acc = 0.845000\n",
      "epoch [1741] L = 38.436295, acc = 0.845000\n",
      "epoch [1742] L = 38.436232, acc = 0.845000\n",
      "epoch [1743] L = 38.436168, acc = 0.845000\n",
      "epoch [1744] L = 38.436105, acc = 0.845000\n",
      "epoch [1745] L = 38.436041, acc = 0.845000\n",
      "epoch [1746] L = 38.435978, acc = 0.845000\n",
      "epoch [1747] L = 38.435915, acc = 0.845000\n",
      "epoch [1748] L = 38.435852, acc = 0.845000\n",
      "epoch [1749] L = 38.435788, acc = 0.845000\n",
      "epoch [1750] L = 38.435725, acc = 0.845000\n",
      "epoch [1751] L = 38.435662, acc = 0.845000\n",
      "epoch [1752] L = 38.435599, acc = 0.845000\n",
      "epoch [1753] L = 38.435536, acc = 0.845000\n",
      "epoch [1754] L = 38.435473, acc = 0.845000\n",
      "epoch [1755] L = 38.435409, acc = 0.845000\n",
      "epoch [1756] L = 38.435346, acc = 0.845000\n",
      "epoch [1757] L = 38.435284, acc = 0.845000\n",
      "epoch [1758] L = 38.435221, acc = 0.845000\n",
      "epoch [1759] L = 38.435158, acc = 0.845000\n",
      "epoch [1760] L = 38.435095, acc = 0.845000\n",
      "epoch [1761] L = 38.435032, acc = 0.845000\n",
      "epoch [1762] L = 38.434969, acc = 0.845000\n",
      "epoch [1763] L = 38.434906, acc = 0.845000\n",
      "epoch [1764] L = 38.434844, acc = 0.845000\n",
      "epoch [1765] L = 38.434781, acc = 0.845000\n",
      "epoch [1766] L = 38.434718, acc = 0.845000\n",
      "epoch [1767] L = 38.434656, acc = 0.845000\n",
      "epoch [1768] L = 38.434593, acc = 0.845000\n",
      "epoch [1769] L = 38.434531, acc = 0.845000\n",
      "epoch [1770] L = 38.434468, acc = 0.845000\n",
      "epoch [1771] L = 38.434406, acc = 0.845000\n",
      "epoch [1772] L = 38.434343, acc = 0.845000\n",
      "epoch [1773] L = 38.434281, acc = 0.845000\n",
      "epoch [1774] L = 38.434218, acc = 0.845000\n",
      "epoch [1775] L = 38.434156, acc = 0.845000\n",
      "epoch [1776] L = 38.434094, acc = 0.845000\n",
      "epoch [1777] L = 38.434031, acc = 0.845000\n",
      "epoch [1778] L = 38.433969, acc = 0.845000\n",
      "epoch [1779] L = 38.433907, acc = 0.845000\n",
      "epoch [1780] L = 38.433845, acc = 0.845000\n",
      "epoch [1781] L = 38.433782, acc = 0.845000\n",
      "epoch [1782] L = 38.433720, acc = 0.845000\n",
      "epoch [1783] L = 38.433658, acc = 0.845000\n",
      "epoch [1784] L = 38.433596, acc = 0.845000\n",
      "epoch [1785] L = 38.433534, acc = 0.845000\n",
      "epoch [1786] L = 38.433472, acc = 0.845000\n",
      "epoch [1787] L = 38.433410, acc = 0.845000\n",
      "epoch [1788] L = 38.433348, acc = 0.845000\n",
      "epoch [1789] L = 38.433286, acc = 0.845000\n",
      "epoch [1790] L = 38.433225, acc = 0.845000\n",
      "epoch [1791] L = 38.433163, acc = 0.845000\n",
      "epoch [1792] L = 38.433101, acc = 0.845000\n",
      "epoch [1793] L = 38.433039, acc = 0.845000\n",
      "epoch [1794] L = 38.432978, acc = 0.845000\n",
      "epoch [1795] L = 38.432916, acc = 0.845000\n",
      "epoch [1796] L = 38.432854, acc = 0.845000\n",
      "epoch [1797] L = 38.432793, acc = 0.845000\n",
      "epoch [1798] L = 38.432731, acc = 0.845000\n",
      "epoch [1799] L = 38.432669, acc = 0.845000\n",
      "epoch [1800] L = 38.432608, acc = 0.845000\n",
      "epoch [1801] L = 38.432546, acc = 0.845000\n",
      "epoch [1802] L = 38.432485, acc = 0.845000\n",
      "epoch [1803] L = 38.432423, acc = 0.845000\n",
      "epoch [1804] L = 38.432362, acc = 0.845000\n",
      "epoch [1805] L = 38.432301, acc = 0.845000\n",
      "epoch [1806] L = 38.432239, acc = 0.845000\n",
      "epoch [1807] L = 38.432178, acc = 0.845000\n",
      "epoch [1808] L = 38.432117, acc = 0.845000\n",
      "epoch [1809] L = 38.432056, acc = 0.845000\n",
      "epoch [1810] L = 38.431994, acc = 0.845000\n",
      "epoch [1811] L = 38.431933, acc = 0.845000\n",
      "epoch [1812] L = 38.431872, acc = 0.845000\n",
      "epoch [1813] L = 38.431811, acc = 0.845000\n",
      "epoch [1814] L = 38.431750, acc = 0.845000\n",
      "epoch [1815] L = 38.431689, acc = 0.845000\n",
      "epoch [1816] L = 38.431628, acc = 0.845000\n",
      "epoch [1817] L = 38.431567, acc = 0.845000\n",
      "epoch [1818] L = 38.431506, acc = 0.845000\n",
      "epoch [1819] L = 38.431445, acc = 0.845000\n",
      "epoch [1820] L = 38.431384, acc = 0.845000\n",
      "epoch [1821] L = 38.431323, acc = 0.845000\n",
      "epoch [1822] L = 38.431263, acc = 0.845000\n",
      "epoch [1823] L = 38.431202, acc = 0.845000\n",
      "epoch [1824] L = 38.431141, acc = 0.845000\n",
      "epoch [1825] L = 38.431080, acc = 0.845000\n",
      "epoch [1826] L = 38.431020, acc = 0.845000\n",
      "epoch [1827] L = 38.430959, acc = 0.845000\n",
      "epoch [1828] L = 38.430898, acc = 0.845000\n",
      "epoch [1829] L = 38.430838, acc = 0.845000\n",
      "epoch [1830] L = 38.430777, acc = 0.845000\n",
      "epoch [1831] L = 38.430717, acc = 0.845000\n",
      "epoch [1832] L = 38.430656, acc = 0.845000\n",
      "epoch [1833] L = 38.430596, acc = 0.845000\n",
      "epoch [1834] L = 38.430535, acc = 0.845000\n",
      "epoch [1835] L = 38.430475, acc = 0.845000\n",
      "epoch [1836] L = 38.430414, acc = 0.845000\n",
      "epoch [1837] L = 38.430354, acc = 0.845000\n",
      "epoch [1838] L = 38.430294, acc = 0.845000\n",
      "epoch [1839] L = 38.430234, acc = 0.845000\n",
      "epoch [1840] L = 38.430173, acc = 0.845000\n",
      "epoch [1841] L = 38.430113, acc = 0.845000\n",
      "epoch [1842] L = 38.430053, acc = 0.845000\n",
      "epoch [1843] L = 38.429993, acc = 0.845000\n",
      "epoch [1844] L = 38.429933, acc = 0.845000\n",
      "epoch [1845] L = 38.429873, acc = 0.845000\n",
      "epoch [1846] L = 38.429812, acc = 0.845000\n",
      "epoch [1847] L = 38.429752, acc = 0.845000\n",
      "epoch [1848] L = 38.429692, acc = 0.845000\n",
      "epoch [1849] L = 38.429632, acc = 0.845000\n",
      "epoch [1850] L = 38.429573, acc = 0.845000\n",
      "epoch [1851] L = 38.429513, acc = 0.845000\n",
      "epoch [1852] L = 38.429453, acc = 0.845000\n",
      "epoch [1853] L = 38.429393, acc = 0.845000\n",
      "epoch [1854] L = 38.429333, acc = 0.845000\n",
      "epoch [1855] L = 38.429273, acc = 0.845000\n",
      "epoch [1856] L = 38.429214, acc = 0.845000\n",
      "epoch [1857] L = 38.429154, acc = 0.845000\n",
      "epoch [1858] L = 38.429094, acc = 0.845000\n",
      "epoch [1859] L = 38.429034, acc = 0.845000\n",
      "epoch [1860] L = 38.428975, acc = 0.845000\n",
      "epoch [1861] L = 38.428915, acc = 0.845000\n",
      "epoch [1862] L = 38.428856, acc = 0.845000\n",
      "epoch [1863] L = 38.428796, acc = 0.845000\n",
      "epoch [1864] L = 38.428737, acc = 0.845000\n",
      "epoch [1865] L = 38.428677, acc = 0.845000\n",
      "epoch [1866] L = 38.428618, acc = 0.845000\n",
      "epoch [1867] L = 38.428558, acc = 0.845000\n",
      "epoch [1868] L = 38.428499, acc = 0.845000\n",
      "epoch [1869] L = 38.428439, acc = 0.845000\n",
      "epoch [1870] L = 38.428380, acc = 0.845000\n",
      "epoch [1871] L = 38.428321, acc = 0.845000\n",
      "epoch [1872] L = 38.428262, acc = 0.845000\n",
      "epoch [1873] L = 38.428202, acc = 0.845000\n",
      "epoch [1874] L = 38.428143, acc = 0.845000\n",
      "epoch [1875] L = 38.428084, acc = 0.845000\n",
      "epoch [1876] L = 38.428025, acc = 0.845000\n",
      "epoch [1877] L = 38.427966, acc = 0.845000\n",
      "epoch [1878] L = 38.427907, acc = 0.845000\n",
      "epoch [1879] L = 38.427848, acc = 0.845000\n",
      "epoch [1880] L = 38.427788, acc = 0.845000\n",
      "epoch [1881] L = 38.427729, acc = 0.845000\n",
      "epoch [1882] L = 38.427671, acc = 0.845000\n",
      "epoch [1883] L = 38.427612, acc = 0.845000\n",
      "epoch [1884] L = 38.427553, acc = 0.845000\n",
      "epoch [1885] L = 38.427494, acc = 0.845000\n",
      "epoch [1886] L = 38.427435, acc = 0.845000\n",
      "epoch [1887] L = 38.427376, acc = 0.845000\n",
      "epoch [1888] L = 38.427317, acc = 0.845000\n",
      "epoch [1889] L = 38.427259, acc = 0.845000\n",
      "epoch [1890] L = 38.427200, acc = 0.845000\n",
      "epoch [1891] L = 38.427141, acc = 0.845000\n",
      "epoch [1892] L = 38.427082, acc = 0.845000\n",
      "epoch [1893] L = 38.427024, acc = 0.845000\n",
      "epoch [1894] L = 38.426965, acc = 0.845000\n",
      "epoch [1895] L = 38.426907, acc = 0.845000\n",
      "epoch [1896] L = 38.426848, acc = 0.845000\n",
      "epoch [1897] L = 38.426789, acc = 0.845000\n",
      "epoch [1898] L = 38.426731, acc = 0.845000\n",
      "epoch [1899] L = 38.426672, acc = 0.845000\n",
      "epoch [1900] L = 38.426614, acc = 0.845000\n",
      "epoch [1901] L = 38.426556, acc = 0.845000\n",
      "epoch [1902] L = 38.426497, acc = 0.845000\n",
      "epoch [1903] L = 38.426439, acc = 0.845000\n",
      "epoch [1904] L = 38.426380, acc = 0.845000\n",
      "epoch [1905] L = 38.426322, acc = 0.845000\n",
      "epoch [1906] L = 38.426264, acc = 0.845000\n",
      "epoch [1907] L = 38.426206, acc = 0.845000\n",
      "epoch [1908] L = 38.426147, acc = 0.845000\n",
      "epoch [1909] L = 38.426089, acc = 0.845000\n",
      "epoch [1910] L = 38.426031, acc = 0.845000\n",
      "epoch [1911] L = 38.425973, acc = 0.845000\n",
      "epoch [1912] L = 38.425915, acc = 0.845000\n",
      "epoch [1913] L = 38.425857, acc = 0.845000\n",
      "epoch [1914] L = 38.425799, acc = 0.845000\n",
      "epoch [1915] L = 38.425741, acc = 0.845000\n",
      "epoch [1916] L = 38.425683, acc = 0.845000\n",
      "epoch [1917] L = 38.425625, acc = 0.845000\n",
      "epoch [1918] L = 38.425567, acc = 0.845000\n",
      "epoch [1919] L = 38.425509, acc = 0.845000\n",
      "epoch [1920] L = 38.425451, acc = 0.845000\n",
      "epoch [1921] L = 38.425393, acc = 0.845000\n",
      "epoch [1922] L = 38.425335, acc = 0.845000\n",
      "epoch [1923] L = 38.425278, acc = 0.845000\n",
      "epoch [1924] L = 38.425220, acc = 0.845000\n",
      "epoch [1925] L = 38.425162, acc = 0.845000\n",
      "epoch [1926] L = 38.425104, acc = 0.845000\n",
      "epoch [1927] L = 38.425047, acc = 0.845000\n",
      "epoch [1928] L = 38.424989, acc = 0.845000\n",
      "epoch [1929] L = 38.424931, acc = 0.845000\n",
      "epoch [1930] L = 38.424874, acc = 0.845000\n",
      "epoch [1931] L = 38.424816, acc = 0.845000\n",
      "epoch [1932] L = 38.424759, acc = 0.845000\n",
      "epoch [1933] L = 38.424701, acc = 0.845000\n",
      "epoch [1934] L = 38.424644, acc = 0.845000\n",
      "epoch [1935] L = 38.424586, acc = 0.845000\n",
      "epoch [1936] L = 38.424529, acc = 0.845000\n",
      "epoch [1937] L = 38.424471, acc = 0.845000\n",
      "epoch [1938] L = 38.424414, acc = 0.845000\n",
      "epoch [1939] L = 38.424357, acc = 0.845000\n",
      "epoch [1940] L = 38.424299, acc = 0.845000\n",
      "epoch [1941] L = 38.424242, acc = 0.845000\n",
      "epoch [1942] L = 38.424185, acc = 0.845000\n",
      "epoch [1943] L = 38.424128, acc = 0.845000\n",
      "epoch [1944] L = 38.424070, acc = 0.845000\n",
      "epoch [1945] L = 38.424013, acc = 0.845000\n",
      "epoch [1946] L = 38.423956, acc = 0.845000\n",
      "epoch [1947] L = 38.423899, acc = 0.845000\n",
      "epoch [1948] L = 38.423842, acc = 0.845000\n",
      "epoch [1949] L = 38.423785, acc = 0.845000\n",
      "epoch [1950] L = 38.423728, acc = 0.845000\n",
      "epoch [1951] L = 38.423671, acc = 0.845000\n",
      "epoch [1952] L = 38.423614, acc = 0.845000\n",
      "epoch [1953] L = 38.423557, acc = 0.845000\n",
      "epoch [1954] L = 38.423500, acc = 0.845000\n",
      "epoch [1955] L = 38.423443, acc = 0.845000\n",
      "epoch [1956] L = 38.423386, acc = 0.845000\n",
      "epoch [1957] L = 38.423329, acc = 0.845000\n",
      "epoch [1958] L = 38.423272, acc = 0.845000\n",
      "epoch [1959] L = 38.423215, acc = 0.845000\n",
      "epoch [1960] L = 38.423159, acc = 0.845000\n",
      "epoch [1961] L = 38.423102, acc = 0.845000\n",
      "epoch [1962] L = 38.423045, acc = 0.845000\n",
      "epoch [1963] L = 38.422989, acc = 0.845000\n",
      "epoch [1964] L = 38.422932, acc = 0.845000\n",
      "epoch [1965] L = 38.422875, acc = 0.845000\n",
      "epoch [1966] L = 38.422819, acc = 0.845000\n",
      "epoch [1967] L = 38.422762, acc = 0.845000\n",
      "epoch [1968] L = 38.422705, acc = 0.845000\n",
      "epoch [1969] L = 38.422649, acc = 0.845000\n",
      "epoch [1970] L = 38.422592, acc = 0.845000\n",
      "epoch [1971] L = 38.422536, acc = 0.845000\n",
      "epoch [1972] L = 38.422479, acc = 0.845000\n",
      "epoch [1973] L = 38.422423, acc = 0.845000\n",
      "epoch [1974] L = 38.422367, acc = 0.845000\n",
      "epoch [1975] L = 38.422310, acc = 0.845000\n",
      "epoch [1976] L = 38.422254, acc = 0.845000\n",
      "epoch [1977] L = 38.422198, acc = 0.845000\n",
      "epoch [1978] L = 38.422141, acc = 0.845000\n",
      "epoch [1979] L = 38.422085, acc = 0.845000\n",
      "epoch [1980] L = 38.422029, acc = 0.845000\n",
      "epoch [1981] L = 38.421973, acc = 0.845000\n",
      "epoch [1982] L = 38.421916, acc = 0.845000\n",
      "epoch [1983] L = 38.421860, acc = 0.845000\n",
      "epoch [1984] L = 38.421804, acc = 0.845000\n",
      "epoch [1985] L = 38.421748, acc = 0.845000\n",
      "epoch [1986] L = 38.421692, acc = 0.845000\n",
      "epoch [1987] L = 38.421636, acc = 0.845000\n",
      "epoch [1988] L = 38.421580, acc = 0.845000\n",
      "epoch [1989] L = 38.421524, acc = 0.845000\n",
      "epoch [1990] L = 38.421468, acc = 0.845000\n",
      "epoch [1991] L = 38.421412, acc = 0.845000\n",
      "epoch [1992] L = 38.421356, acc = 0.845000\n",
      "epoch [1993] L = 38.421300, acc = 0.845000\n",
      "epoch [1994] L = 38.421244, acc = 0.845000\n",
      "epoch [1995] L = 38.421188, acc = 0.845000\n",
      "epoch [1996] L = 38.421132, acc = 0.845000\n",
      "epoch [1997] L = 38.421077, acc = 0.845000\n",
      "epoch [1998] L = 38.421021, acc = 0.845000\n",
      "epoch [1999] L = 38.420965, acc = 0.845000\n"
     ]
    }
   ],
   "source": [
    "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": 5,
   "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": [
    "# 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": [
    "## 如何使用类的方法封装多层神经网络?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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, 4, 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": 9,
   "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": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "L = 107.040493, acc = 0.500000\n",
      "L = 102.045362, acc = 0.500000\n",
      "L = 97.906443, acc = 0.500000\n",
      "L = 94.853004, acc = 0.500000\n",
      "L = 92.680087, acc = 0.515000\n",
      "L = 91.034467, acc = 0.675000\n",
      "L = 89.646974, acc = 0.740000\n",
      "L = 88.369257, acc = 0.760000\n",
      "L = 87.131907, acc = 0.800000\n",
      "L = 85.905402, acc = 0.790000\n",
      "L = 84.678282, acc = 0.795000\n",
      "L = 83.446916, acc = 0.800000\n",
      "L = 82.211090, acc = 0.795000\n",
      "L = 80.972156, acc = 0.795000\n",
      "L = 79.732278, acc = 0.795000\n",
      "L = 78.494093, acc = 0.810000\n",
      "L = 77.260550, acc = 0.815000\n",
      "L = 76.034802, acc = 0.805000\n",
      "L = 74.820121, acc = 0.805000\n",
      "L = 73.619820, acc = 0.805000\n",
      "L = 72.437170, acc = 0.810000\n",
      "L = 71.275332, acc = 0.810000\n",
      "L = 70.137285, acc = 0.810000\n",
      "L = 69.025762, acc = 0.810000\n",
      "L = 67.943204, acc = 0.810000\n",
      "L = 66.891717, acc = 0.810000\n",
      "L = 65.873045, acc = 0.810000\n",
      "L = 64.888558, acc = 0.805000\n",
      "L = 63.939247, acc = 0.805000\n",
      "L = 63.025738, acc = 0.800000\n",
      "L = 62.148308, acc = 0.800000\n",
      "L = 61.306913, acc = 0.800000\n",
      "L = 60.501221, acc = 0.805000\n",
      "L = 59.730650, acc = 0.805000\n",
      "L = 58.994403, acc = 0.805000\n",
      "L = 58.291506, acc = 0.805000\n",
      "L = 57.620848, acc = 0.805000\n",
      "L = 56.981212, acc = 0.805000\n",
      "L = 56.371308, acc = 0.805000\n",
      "L = 55.789797, acc = 0.805000\n",
      "L = 55.235320, acc = 0.805000\n",
      "L = 54.706516, acc = 0.805000\n",
      "L = 54.202038, acc = 0.810000\n",
      "L = 53.720565, acc = 0.815000\n",
      "L = 53.260818, acc = 0.820000\n",
      "L = 52.821562, acc = 0.820000\n",
      "L = 52.401615, acc = 0.820000\n",
      "L = 51.999850, acc = 0.820000\n",
      "L = 51.615198, acc = 0.820000\n",
      "L = 51.246651, acc = 0.820000\n",
      "L = 50.893258, acc = 0.820000\n",
      "L = 50.554124, acc = 0.820000\n",
      "L = 50.228415, acc = 0.820000\n",
      "L = 49.915347, acc = 0.820000\n",
      "L = 49.614189, acc = 0.820000\n",
      "L = 49.324260, acc = 0.820000\n",
      "L = 49.044927, acc = 0.820000\n",
      "L = 48.775597, acc = 0.820000\n",
      "L = 48.515723, acc = 0.820000\n",
      "L = 48.264794, acc = 0.820000\n",
      "L = 48.022334, acc = 0.820000\n",
      "L = 47.787904, acc = 0.820000\n",
      "L = 47.561092, acc = 0.820000\n",
      "L = 47.341518, acc = 0.820000\n",
      "L = 47.128828, acc = 0.820000\n",
      "L = 46.922692, acc = 0.820000\n",
      "L = 46.722802, acc = 0.820000\n",
      "L = 46.528875, acc = 0.820000\n",
      "L = 46.340642, acc = 0.825000\n",
      "L = 46.157856, acc = 0.825000\n",
      "L = 45.980285, acc = 0.825000\n",
      "L = 45.807711, acc = 0.825000\n",
      "L = 45.639934, acc = 0.825000\n",
      "L = 45.476762, acc = 0.825000\n",
      "L = 45.318017, acc = 0.825000\n",
      "L = 45.163533, acc = 0.825000\n",
      "L = 45.013153, acc = 0.825000\n",
      "L = 44.866728, acc = 0.825000\n",
      "L = 44.724120, acc = 0.825000\n",
      "L = 44.585197, acc = 0.825000\n",
      "L = 44.449835, acc = 0.825000\n",
      "L = 44.317917, acc = 0.825000\n",
      "L = 44.189331, acc = 0.825000\n",
      "L = 44.063971, acc = 0.825000\n",
      "L = 43.941737, acc = 0.825000\n",
      "L = 43.822534, acc = 0.830000\n",
      "L = 43.706270, acc = 0.835000\n",
      "L = 43.592859, acc = 0.835000\n",
      "L = 43.482217, acc = 0.835000\n",
      "L = 43.374264, acc = 0.835000\n",
      "L = 43.268926, acc = 0.835000\n",
      "L = 43.166127, acc = 0.835000\n",
      "L = 43.065799, acc = 0.840000\n",
      "L = 42.967874, acc = 0.835000\n",
      "L = 42.872286, acc = 0.840000\n",
      "L = 42.778973, acc = 0.840000\n",
      "L = 42.687875, acc = 0.840000\n",
      "L = 42.598932, acc = 0.840000\n",
      "L = 42.512089, acc = 0.840000\n",
      "L = 42.427290, acc = 0.840000\n",
      "L = 42.344484, acc = 0.840000\n",
      "L = 42.263618, acc = 0.840000\n",
      "L = 42.184643, acc = 0.840000\n",
      "L = 42.107510, acc = 0.840000\n",
      "L = 42.032174, acc = 0.840000\n",
      "L = 41.958587, acc = 0.840000\n",
      "L = 41.886707, acc = 0.840000\n",
      "L = 41.816490, acc = 0.840000\n",
      "L = 41.747895, acc = 0.840000\n",
      "L = 41.680880, acc = 0.835000\n",
      "L = 41.615406, acc = 0.835000\n",
      "L = 41.551435, acc = 0.835000\n",
      "L = 41.488930, acc = 0.835000\n",
      "L = 41.427852, acc = 0.835000\n",
      "L = 41.368168, acc = 0.835000\n",
      "L = 41.309842, acc = 0.835000\n",
      "L = 41.252841, acc = 0.835000\n",
      "L = 41.197132, acc = 0.835000\n",
      "L = 41.142682, acc = 0.835000\n",
      "L = 41.089460, acc = 0.835000\n",
      "L = 41.037437, acc = 0.835000\n",
      "L = 40.986581, acc = 0.835000\n",
      "L = 40.936865, acc = 0.840000\n",
      "L = 40.888260, acc = 0.840000\n",
      "L = 40.840738, acc = 0.840000\n",
      "L = 40.794273, acc = 0.840000\n",
      "L = 40.748839, acc = 0.840000\n",
      "L = 40.704410, acc = 0.840000\n",
      "L = 40.660961, acc = 0.840000\n",
      "L = 40.618468, acc = 0.840000\n",
      "L = 40.576908, acc = 0.840000\n",
      "L = 40.536257, acc = 0.845000\n",
      "L = 40.496494, acc = 0.845000\n",
      "L = 40.457596, acc = 0.845000\n",
      "L = 40.419541, acc = 0.850000\n",
      "L = 40.382310, acc = 0.850000\n",
      "L = 40.345882, acc = 0.850000\n",
      "L = 40.310238, acc = 0.855000\n",
      "L = 40.275357, acc = 0.855000\n",
      "L = 40.241222, acc = 0.855000\n",
      "L = 40.207815, acc = 0.855000\n",
      "L = 40.175117, acc = 0.855000\n",
      "L = 40.143111, acc = 0.855000\n",
      "L = 40.111781, acc = 0.855000\n",
      "L = 40.081110, acc = 0.855000\n",
      "L = 40.051081, acc = 0.855000\n",
      "L = 40.021681, acc = 0.855000\n",
      "L = 39.992893, acc = 0.855000\n",
      "L = 39.964702, acc = 0.855000\n",
      "L = 39.937094, acc = 0.855000\n",
      "L = 39.910056, acc = 0.855000\n",
      "L = 39.883573, acc = 0.855000\n",
      "L = 39.857633, acc = 0.855000\n",
      "L = 39.832222, acc = 0.855000\n",
      "L = 39.807327, acc = 0.855000\n",
      "L = 39.782937, acc = 0.855000\n",
      "L = 39.759039, acc = 0.855000\n",
      "L = 39.735623, acc = 0.855000\n",
      "L = 39.712675, acc = 0.855000\n",
      "L = 39.690186, acc = 0.855000\n",
      "L = 39.668144, acc = 0.855000\n",
      "L = 39.646540, acc = 0.855000\n",
      "L = 39.625362, acc = 0.855000\n",
      "L = 39.604602, acc = 0.855000\n",
      "L = 39.584248, acc = 0.855000\n",
      "L = 39.564293, acc = 0.855000\n",
      "L = 39.544726, acc = 0.855000\n",
      "L = 39.525538, acc = 0.855000\n",
      "L = 39.506721, acc = 0.855000\n",
      "L = 39.488267, acc = 0.855000\n",
      "L = 39.470166, acc = 0.855000\n",
      "L = 39.452412, acc = 0.855000\n",
      "L = 39.434995, acc = 0.855000\n",
      "L = 39.417909, acc = 0.855000\n",
      "L = 39.401146, acc = 0.855000\n",
      "L = 39.384698, acc = 0.855000\n",
      "L = 39.368558, acc = 0.855000\n",
      "L = 39.352721, acc = 0.855000\n",
      "L = 39.337178, acc = 0.855000\n",
      "L = 39.321923, acc = 0.855000\n",
      "L = 39.306950, acc = 0.855000\n",
      "L = 39.292252, acc = 0.855000\n",
      "L = 39.277824, acc = 0.855000\n",
      "L = 39.263659, acc = 0.855000\n",
      "L = 39.249751, acc = 0.855000\n",
      "L = 39.236096, acc = 0.855000\n",
      "L = 39.222687, acc = 0.855000\n",
      "L = 39.209520, acc = 0.860000\n",
      "L = 39.196588, acc = 0.860000\n",
      "L = 39.183886, acc = 0.860000\n",
      "L = 39.171411, acc = 0.860000\n",
      "L = 39.159156, acc = 0.860000\n",
      "L = 39.147117, acc = 0.860000\n",
      "L = 39.135290, acc = 0.860000\n",
      "L = 39.123669, acc = 0.860000\n",
      "L = 39.112251, acc = 0.860000\n",
      "L = 39.101031, acc = 0.860000\n",
      "L = 39.090005, acc = 0.860000\n",
      "L = 39.079169, acc = 0.860000\n",
      "L = 39.068519, acc = 0.860000\n",
      "L = 39.058050, acc = 0.865000\n",
      "L = 39.047760, acc = 0.865000\n",
      "L = 39.037644, acc = 0.865000\n",
      "L = 39.027699, acc = 0.865000\n",
      "L = 39.017920, acc = 0.865000\n",
      "L = 39.008306, acc = 0.865000\n",
      "L = 38.998852, acc = 0.865000\n",
      "L = 38.989555, acc = 0.865000\n",
      "L = 38.980411, acc = 0.865000\n",
      "L = 38.971418, acc = 0.865000\n",
      "L = 38.962573, acc = 0.865000\n",
      "L = 38.953872, acc = 0.865000\n",
      "L = 38.945312, acc = 0.865000\n",
      "L = 38.936892, acc = 0.865000\n",
      "L = 38.928607, acc = 0.865000\n",
      "L = 38.920455, acc = 0.865000\n",
      "L = 38.912433, acc = 0.865000\n",
      "L = 38.904539, acc = 0.865000\n",
      "L = 38.896771, acc = 0.865000\n",
      "L = 38.889125, acc = 0.865000\n",
      "L = 38.881600, acc = 0.865000\n",
      "L = 38.874192, acc = 0.865000\n",
      "L = 38.866900, acc = 0.865000\n",
      "L = 38.859720, acc = 0.865000\n",
      "L = 38.852652, acc = 0.865000\n",
      "L = 38.845693, acc = 0.865000\n",
      "L = 38.838840, acc = 0.865000\n",
      "L = 38.832092, acc = 0.865000\n",
      "L = 38.825446, acc = 0.865000\n",
      "L = 38.818901, acc = 0.865000\n",
      "L = 38.812455, acc = 0.865000\n",
      "L = 38.806104, acc = 0.865000\n",
      "L = 38.799849, acc = 0.865000\n",
      "L = 38.793686, acc = 0.865000\n",
      "L = 38.787615, acc = 0.865000\n",
      "L = 38.781633, acc = 0.865000\n",
      "L = 38.775738, acc = 0.865000\n",
      "L = 38.769930, acc = 0.865000\n",
      "L = 38.764205, acc = 0.865000\n",
      "L = 38.758563, acc = 0.865000\n",
      "L = 38.753003, acc = 0.865000\n",
      "L = 38.747521, acc = 0.865000\n",
      "L = 38.742118, acc = 0.865000\n",
      "L = 38.736791, acc = 0.865000\n",
      "L = 38.731540, acc = 0.865000\n",
      "L = 38.726362, acc = 0.865000\n",
      "L = 38.721256, acc = 0.865000\n",
      "L = 38.716221, acc = 0.865000\n",
      "L = 38.711255, acc = 0.865000\n",
      "L = 38.706358, acc = 0.865000\n",
      "L = 38.701528, acc = 0.865000\n",
      "L = 38.696763, acc = 0.865000\n",
      "L = 38.692063, acc = 0.865000\n",
      "L = 38.687426, acc = 0.865000\n",
      "L = 38.682851, acc = 0.865000\n",
      "L = 38.678337, acc = 0.865000\n",
      "L = 38.673883, acc = 0.865000\n",
      "L = 38.669488, acc = 0.865000\n",
      "L = 38.665150, acc = 0.865000\n",
      "L = 38.660868, acc = 0.865000\n",
      "L = 38.656642, acc = 0.865000\n",
      "L = 38.652470, acc = 0.865000\n",
      "L = 38.648352, acc = 0.865000\n",
      "L = 38.644286, acc = 0.865000\n",
      "L = 38.640272, acc = 0.865000\n",
      "L = 38.636308, acc = 0.865000\n",
      "L = 38.632394, acc = 0.865000\n",
      "L = 38.628528, acc = 0.865000\n",
      "L = 38.624710, acc = 0.865000\n",
      "L = 38.620939, acc = 0.865000\n",
      "L = 38.617214, acc = 0.865000\n",
      "L = 38.613534, acc = 0.865000\n",
      "L = 38.609898, acc = 0.865000\n",
      "L = 38.606306, acc = 0.865000\n",
      "L = 38.602757, acc = 0.865000\n",
      "L = 38.599250, acc = 0.865000\n",
      "L = 38.595783, acc = 0.865000\n",
      "L = 38.592358, acc = 0.865000\n",
      "L = 38.588971, acc = 0.865000\n",
      "L = 38.585624, acc = 0.865000\n",
      "L = 38.582315, acc = 0.865000\n",
      "L = 38.579043, acc = 0.865000\n",
      "L = 38.575809, acc = 0.865000\n",
      "L = 38.572610, acc = 0.865000\n",
      "L = 38.569447, acc = 0.865000\n",
      "L = 38.566319, acc = 0.865000\n",
      "L = 38.563225, acc = 0.865000\n",
      "L = 38.560165, acc = 0.865000\n",
      "L = 38.557137, acc = 0.865000\n",
      "L = 38.554142, acc = 0.865000\n",
      "L = 38.551178, acc = 0.865000\n",
      "L = 38.548246, acc = 0.865000\n",
      "L = 38.545344, acc = 0.865000\n",
      "L = 38.542473, acc = 0.865000\n",
      "L = 38.539631, acc = 0.865000\n",
      "L = 38.536817, acc = 0.865000\n",
      "L = 38.534032, acc = 0.865000\n",
      "L = 38.531275, acc = 0.865000\n",
      "L = 38.528546, acc = 0.865000\n",
      "L = 38.525843, acc = 0.865000\n",
      "L = 38.523166, acc = 0.865000\n",
      "L = 38.520515, acc = 0.865000\n",
      "L = 38.517890, acc = 0.865000\n",
      "L = 38.515289, acc = 0.865000\n",
      "L = 38.512713, acc = 0.865000\n",
      "L = 38.510161, acc = 0.865000\n",
      "L = 38.507632, acc = 0.865000\n",
      "L = 38.505126, acc = 0.865000\n",
      "L = 38.502643, acc = 0.865000\n",
      "L = 38.500182, acc = 0.865000\n",
      "L = 38.497743, acc = 0.865000\n",
      "L = 38.495325, acc = 0.865000\n",
      "L = 38.492928, acc = 0.865000\n",
      "L = 38.490552, acc = 0.865000\n",
      "L = 38.488195, acc = 0.865000\n",
      "L = 38.485859, acc = 0.865000\n",
      "L = 38.483541, acc = 0.865000\n",
      "L = 38.481243, acc = 0.865000\n",
      "L = 38.478963, acc = 0.865000\n",
      "L = 38.476702, acc = 0.865000\n",
      "L = 38.474458, acc = 0.865000\n",
      "L = 38.472232, acc = 0.865000\n",
      "L = 38.470023, acc = 0.865000\n",
      "L = 38.467830, acc = 0.865000\n",
      "L = 38.465654, acc = 0.865000\n",
      "L = 38.463495, acc = 0.865000\n",
      "L = 38.461351, acc = 0.865000\n",
      "L = 38.459222, acc = 0.865000\n",
      "L = 38.457109, acc = 0.865000\n",
      "L = 38.455010, acc = 0.865000\n",
      "L = 38.452926, acc = 0.865000\n",
      "L = 38.450856, acc = 0.865000\n",
      "L = 38.448800, acc = 0.865000\n",
      "L = 38.446758, acc = 0.865000\n",
      "L = 38.444729, acc = 0.865000\n",
      "L = 38.442713, acc = 0.865000\n",
      "L = 38.440709, acc = 0.865000\n",
      "L = 38.438718, acc = 0.865000\n",
      "L = 38.436739, acc = 0.865000\n",
      "L = 38.434772, acc = 0.865000\n",
      "L = 38.432817, acc = 0.865000\n",
      "L = 38.430873, acc = 0.865000\n",
      "L = 38.428940, acc = 0.865000\n",
      "L = 38.427018, acc = 0.865000\n",
      "L = 38.425106, acc = 0.865000\n",
      "L = 38.423205, acc = 0.865000\n",
      "L = 38.421314, acc = 0.865000\n",
      "L = 38.419432, acc = 0.865000\n",
      "L = 38.417561, acc = 0.865000\n",
      "L = 38.415698, acc = 0.865000\n",
      "L = 38.413844, acc = 0.865000\n",
      "L = 38.412000, acc = 0.865000\n",
      "L = 38.410164, acc = 0.865000\n",
      "L = 38.408336, acc = 0.865000\n",
      "L = 38.406516, acc = 0.865000\n",
      "L = 38.404705, acc = 0.865000\n",
      "L = 38.402901, acc = 0.865000\n",
      "L = 38.401104, acc = 0.865000\n",
      "L = 38.399315, acc = 0.865000\n",
      "L = 38.397533, acc = 0.865000\n",
      "L = 38.395757, acc = 0.865000\n",
      "L = 38.393989, acc = 0.865000\n",
      "L = 38.392226, acc = 0.865000\n",
      "L = 38.390470, acc = 0.865000\n",
      "L = 38.388720, acc = 0.865000\n",
      "L = 38.386975, acc = 0.865000\n",
      "L = 38.385237, acc = 0.865000\n",
      "L = 38.383503, acc = 0.865000\n",
      "L = 38.381775, acc = 0.865000\n",
      "L = 38.380052, acc = 0.865000\n",
      "L = 38.378333, acc = 0.865000\n",
      "L = 38.376619, acc = 0.865000\n",
      "L = 38.374910, acc = 0.865000\n",
      "L = 38.373205, acc = 0.865000\n",
      "L = 38.371504, acc = 0.865000\n",
      "L = 38.369807, acc = 0.865000\n",
      "L = 38.368113, acc = 0.865000\n",
      "L = 38.366423, acc = 0.865000\n",
      "L = 38.364737, acc = 0.865000\n",
      "L = 38.363053, acc = 0.865000\n",
      "L = 38.361373, acc = 0.865000\n",
      "L = 38.359695, acc = 0.865000\n",
      "L = 38.358020, acc = 0.865000\n",
      "L = 38.356347, acc = 0.865000\n",
      "L = 38.354677, acc = 0.865000\n",
      "L = 38.353009, acc = 0.865000\n",
      "L = 38.351343, acc = 0.865000\n",
      "L = 38.349678, acc = 0.865000\n",
      "L = 38.348015, acc = 0.865000\n",
      "L = 38.346354, acc = 0.865000\n",
      "L = 38.344694, acc = 0.865000\n",
      "L = 38.343035, acc = 0.865000\n",
      "L = 38.341377, acc = 0.865000\n",
      "L = 38.339720, acc = 0.865000\n",
      "L = 38.338064, acc = 0.865000\n",
      "L = 38.336408, acc = 0.865000\n",
      "L = 38.334752, acc = 0.865000\n",
      "L = 38.333097, acc = 0.865000\n",
      "L = 38.331442, acc = 0.865000\n",
      "L = 38.329786, acc = 0.865000\n",
      "L = 38.328131, acc = 0.865000\n",
      "L = 38.326475, acc = 0.865000\n",
      "L = 38.324818, acc = 0.865000\n",
      "L = 38.323161, acc = 0.865000\n",
      "L = 38.321502, acc = 0.865000\n",
      "L = 38.319843, acc = 0.865000\n",
      "L = 38.318183, acc = 0.865000\n",
      "L = 38.316521, acc = 0.865000\n",
      "L = 38.314858, acc = 0.865000\n",
      "L = 38.313193, acc = 0.865000\n",
      "L = 38.311527, acc = 0.865000\n",
      "L = 38.309858, acc = 0.865000\n",
      "L = 38.308188, acc = 0.865000\n",
      "L = 38.306515, acc = 0.865000\n",
      "L = 38.304840, acc = 0.865000\n",
      "L = 38.303163, acc = 0.865000\n",
      "L = 38.301483, acc = 0.865000\n",
      "L = 38.299800, acc = 0.865000\n",
      "L = 38.298115, acc = 0.865000\n",
      "L = 38.296426, acc = 0.865000\n",
      "L = 38.294734, acc = 0.865000\n",
      "L = 38.293039, acc = 0.865000\n",
      "L = 38.291341, acc = 0.865000\n",
      "L = 38.289639, acc = 0.865000\n",
      "L = 38.287933, acc = 0.865000\n",
      "L = 38.286224, acc = 0.865000\n",
      "L = 38.284510, acc = 0.865000\n",
      "L = 38.282792, acc = 0.865000\n",
      "L = 38.281070, acc = 0.865000\n",
      "L = 38.279344, acc = 0.865000\n",
      "L = 38.277613, acc = 0.865000\n",
      "L = 38.275878, acc = 0.865000\n",
      "L = 38.274137, acc = 0.865000\n",
      "L = 38.272392, acc = 0.865000\n",
      "L = 38.270642, acc = 0.865000\n",
      "L = 38.268886, acc = 0.865000\n",
      "L = 38.267126, acc = 0.865000\n",
      "L = 38.265359, acc = 0.865000\n",
      "L = 38.263587, acc = 0.865000\n",
      "L = 38.261810, acc = 0.865000\n",
      "L = 38.260026, acc = 0.865000\n",
      "L = 38.258237, acc = 0.865000\n",
      "L = 38.256441, acc = 0.865000\n",
      "L = 38.254639, acc = 0.865000\n",
      "L = 38.252831, acc = 0.865000\n",
      "L = 38.251016, acc = 0.865000\n",
      "L = 38.249195, acc = 0.865000\n",
      "L = 38.247367, acc = 0.865000\n",
      "L = 38.245532, acc = 0.865000\n",
      "L = 38.243690, acc = 0.865000\n",
      "L = 38.241840, acc = 0.865000\n",
      "L = 38.239984, acc = 0.865000\n",
      "L = 38.238120, acc = 0.865000\n",
      "L = 38.236248, acc = 0.865000\n",
      "L = 38.234369, acc = 0.865000\n",
      "L = 38.232482, acc = 0.865000\n",
      "L = 38.230587, acc = 0.865000\n",
      "L = 38.228683, acc = 0.865000\n",
      "L = 38.226772, acc = 0.865000\n",
      "L = 38.224852, acc = 0.865000\n",
      "L = 38.222924, acc = 0.865000\n",
      "L = 38.220987, acc = 0.865000\n",
      "L = 38.219042, acc = 0.865000\n",
      "L = 38.217087, acc = 0.865000\n",
      "L = 38.215124, acc = 0.865000\n",
      "L = 38.213151, acc = 0.865000\n",
      "L = 38.211170, acc = 0.865000\n",
      "L = 38.209178, acc = 0.865000\n",
      "L = 38.207178, acc = 0.865000\n",
      "L = 38.205167, acc = 0.865000\n",
      "L = 38.203147, acc = 0.865000\n",
      "L = 38.201117, acc = 0.865000\n",
      "L = 38.199078, acc = 0.865000\n",
      "L = 38.197027, acc = 0.865000\n",
      "L = 38.194967, acc = 0.865000\n",
      "L = 38.192896, acc = 0.865000\n",
      "L = 38.190815, acc = 0.865000\n",
      "L = 38.188723, acc = 0.865000\n",
      "L = 38.186620, acc = 0.865000\n",
      "L = 38.184506, acc = 0.865000\n",
      "L = 38.182382, acc = 0.865000\n",
      "L = 38.180246, acc = 0.865000\n",
      "L = 38.178099, acc = 0.865000\n",
      "L = 38.175940, acc = 0.865000\n",
      "L = 38.173770, acc = 0.865000\n",
      "L = 38.171588, acc = 0.865000\n",
      "L = 38.169394, acc = 0.865000\n",
      "L = 38.167189, acc = 0.865000\n",
      "L = 38.164971, acc = 0.865000\n",
      "L = 38.162741, acc = 0.865000\n",
      "L = 38.160499, acc = 0.865000\n",
      "L = 38.158244, acc = 0.865000\n",
      "L = 38.155977, acc = 0.865000\n",
      "L = 38.153697, acc = 0.865000\n",
      "L = 38.151404, acc = 0.865000\n",
      "L = 38.149099, acc = 0.865000\n",
      "L = 38.146780, acc = 0.865000\n",
      "L = 38.144448, acc = 0.865000\n",
      "L = 38.142103, acc = 0.865000\n",
      "L = 38.139744, acc = 0.865000\n",
      "L = 38.137372, acc = 0.865000\n",
      "L = 38.134986, acc = 0.865000\n",
      "L = 38.132586, acc = 0.865000\n",
      "L = 38.130172, acc = 0.865000\n",
      "L = 38.127744, acc = 0.865000\n",
      "L = 38.125302, acc = 0.865000\n",
      "L = 38.122846, acc = 0.865000\n",
      "L = 38.120375, acc = 0.865000\n",
      "L = 38.117890, acc = 0.865000\n",
      "L = 38.115390, acc = 0.865000\n",
      "L = 38.112875, acc = 0.865000\n",
      "L = 38.110345, acc = 0.865000\n",
      "L = 38.107800, acc = 0.865000\n",
      "L = 38.105240, acc = 0.865000\n",
      "L = 38.102664, acc = 0.865000\n",
      "L = 38.100074, acc = 0.865000\n",
      "L = 38.097467, acc = 0.865000\n",
      "L = 38.094845, acc = 0.865000\n",
      "L = 38.092207, acc = 0.865000\n",
      "L = 38.089554, acc = 0.865000\n",
      "L = 38.086884, acc = 0.865000\n",
      "L = 38.084198, acc = 0.865000\n",
      "L = 38.081496, acc = 0.865000\n",
      "L = 38.078777, acc = 0.865000\n",
      "L = 38.076042, acc = 0.865000\n",
      "L = 38.073290, acc = 0.865000\n",
      "L = 38.070522, acc = 0.865000\n",
      "L = 38.067736, acc = 0.865000\n",
      "L = 38.064934, acc = 0.865000\n",
      "L = 38.062114, acc = 0.865000\n",
      "L = 38.059278, acc = 0.865000\n",
      "L = 38.056424, acc = 0.865000\n",
      "L = 38.053552, acc = 0.865000\n",
      "L = 38.050663, acc = 0.865000\n",
      "L = 38.047756, acc = 0.865000\n",
      "L = 38.044832, acc = 0.865000\n",
      "L = 38.041889, acc = 0.865000\n",
      "L = 38.038929, acc = 0.865000\n",
      "L = 38.035950, acc = 0.865000\n",
      "L = 38.032953, acc = 0.865000\n",
      "L = 38.029938, acc = 0.865000\n",
      "L = 38.026904, acc = 0.865000\n",
      "L = 38.023851, acc = 0.865000\n",
      "L = 38.020780, acc = 0.865000\n",
      "L = 38.017690, acc = 0.865000\n",
      "L = 38.014581, acc = 0.865000\n",
      "L = 38.011453, acc = 0.865000\n",
      "L = 38.008306, acc = 0.865000\n",
      "L = 38.005139, acc = 0.865000\n",
      "L = 38.001954, acc = 0.865000\n",
      "L = 37.998748, acc = 0.865000\n",
      "L = 37.995523, acc = 0.865000\n",
      "L = 37.992279, acc = 0.865000\n",
      "L = 37.989014, acc = 0.865000\n",
      "L = 37.985729, acc = 0.865000\n",
      "L = 37.982425, acc = 0.865000\n",
      "L = 37.979100, acc = 0.865000\n",
      "L = 37.975755, acc = 0.865000\n",
      "L = 37.972390, acc = 0.865000\n",
      "L = 37.969004, acc = 0.865000\n",
      "L = 37.965598, acc = 0.865000\n",
      "L = 37.962170, acc = 0.865000\n",
      "L = 37.958722, acc = 0.865000\n",
      "L = 37.955253, acc = 0.865000\n",
      "L = 37.951764, acc = 0.865000\n",
      "L = 37.948253, acc = 0.865000\n",
      "L = 37.944720, acc = 0.865000\n",
      "L = 37.941167, acc = 0.865000\n",
      "L = 37.937591, acc = 0.865000\n",
      "L = 37.933995, acc = 0.865000\n",
      "L = 37.930376, acc = 0.865000\n",
      "L = 37.926736, acc = 0.865000\n",
      "L = 37.923074, acc = 0.865000\n",
      "L = 37.919390, acc = 0.865000\n",
      "L = 37.915684, acc = 0.865000\n",
      "L = 37.911956, acc = 0.865000\n",
      "L = 37.908205, acc = 0.865000\n",
      "L = 37.904432, acc = 0.865000\n",
      "L = 37.900637, acc = 0.865000\n",
      "L = 37.896819, acc = 0.865000\n",
      "L = 37.892978, acc = 0.865000\n",
      "L = 37.889114, acc = 0.865000\n",
      "L = 37.885228, acc = 0.865000\n",
      "L = 37.881318, acc = 0.865000\n",
      "L = 37.877385, acc = 0.865000\n",
      "L = 37.873429, acc = 0.865000\n",
      "L = 37.869449, acc = 0.865000\n",
      "L = 37.865446, acc = 0.865000\n",
      "L = 37.861420, acc = 0.865000\n",
      "L = 37.857369, acc = 0.865000\n",
      "L = 37.853295, acc = 0.865000\n",
      "L = 37.849197, acc = 0.865000\n",
      "L = 37.845075, acc = 0.865000\n",
      "L = 37.840929, acc = 0.865000\n",
      "L = 37.836759, acc = 0.865000\n",
      "L = 37.832564, acc = 0.865000\n",
      "L = 37.828344, acc = 0.865000\n",
      "L = 37.824101, acc = 0.865000\n",
      "L = 37.819832, acc = 0.865000\n",
      "L = 37.815539, acc = 0.865000\n",
      "L = 37.811220, acc = 0.865000\n",
      "L = 37.806877, acc = 0.865000\n",
      "L = 37.802508, acc = 0.865000\n",
      "L = 37.798114, acc = 0.865000\n",
      "L = 37.793695, acc = 0.865000\n",
      "L = 37.789250, acc = 0.865000\n",
      "L = 37.784780, acc = 0.865000\n",
      "L = 37.780283, acc = 0.865000\n",
      "L = 37.775761, acc = 0.865000\n",
      "L = 37.771213, acc = 0.865000\n",
      "L = 37.766639, acc = 0.865000\n",
      "L = 37.762038, acc = 0.865000\n",
      "L = 37.757411, acc = 0.865000\n",
      "L = 37.752758, acc = 0.865000\n",
      "L = 37.748078, acc = 0.865000\n",
      "L = 37.743371, acc = 0.865000\n",
      "L = 37.738637, acc = 0.865000\n",
      "L = 37.733876, acc = 0.865000\n",
      "L = 37.729088, acc = 0.865000\n",
      "L = 37.724272, acc = 0.865000\n",
      "L = 37.719429, acc = 0.865000\n",
      "L = 37.714559, acc = 0.865000\n",
      "L = 37.709660, acc = 0.865000\n",
      "L = 37.704734, acc = 0.865000\n",
      "L = 37.699780, acc = 0.865000\n",
      "L = 37.694797, acc = 0.865000\n",
      "L = 37.689786, acc = 0.865000\n",
      "L = 37.684747, acc = 0.865000\n",
      "L = 37.679679, acc = 0.865000\n",
      "L = 37.674582, acc = 0.865000\n",
      "L = 37.669456, acc = 0.865000\n",
      "L = 37.664301, acc = 0.865000\n",
      "L = 37.659117, acc = 0.865000\n",
      "L = 37.653903, acc = 0.865000\n",
      "L = 37.648659, acc = 0.865000\n",
      "L = 37.643386, acc = 0.865000\n",
      "L = 37.638083, acc = 0.865000\n",
      "L = 37.632749, acc = 0.865000\n",
      "L = 37.627386, acc = 0.865000\n",
      "L = 37.621991, acc = 0.865000\n",
      "L = 37.616566, acc = 0.865000\n",
      "L = 37.611111, acc = 0.865000\n",
      "L = 37.605624, acc = 0.865000\n",
      "L = 37.600106, acc = 0.865000\n",
      "L = 37.594556, acc = 0.865000\n",
      "L = 37.588975, acc = 0.865000\n",
      "L = 37.583361, acc = 0.865000\n",
      "L = 37.577716, acc = 0.865000\n",
      "L = 37.572039, acc = 0.865000\n",
      "L = 37.566329, acc = 0.865000\n",
      "L = 37.560586, acc = 0.865000\n",
      "L = 37.554811, acc = 0.865000\n",
      "L = 37.549002, acc = 0.865000\n",
      "L = 37.543160, acc = 0.865000\n",
      "L = 37.537284, acc = 0.865000\n",
      "L = 37.531375, acc = 0.865000\n",
      "L = 37.525432, acc = 0.865000\n",
      "L = 37.519454, acc = 0.865000\n",
      "L = 37.513442, acc = 0.865000\n",
      "L = 37.507396, acc = 0.870000\n",
      "L = 37.501314, acc = 0.870000\n",
      "L = 37.495197, acc = 0.870000\n",
      "L = 37.489045, acc = 0.870000\n",
      "L = 37.482856, acc = 0.870000\n",
      "L = 37.476632, acc = 0.870000\n",
      "L = 37.470372, acc = 0.870000\n",
      "L = 37.464075, acc = 0.870000\n",
      "L = 37.457742, acc = 0.870000\n",
      "L = 37.451371, acc = 0.870000\n",
      "L = 37.444964, acc = 0.870000\n",
      "L = 37.438518, acc = 0.870000\n",
      "L = 37.432035, acc = 0.870000\n",
      "L = 37.425514, acc = 0.870000\n",
      "L = 37.418954, acc = 0.870000\n",
      "L = 37.412355, acc = 0.870000\n",
      "L = 37.405718, acc = 0.870000\n",
      "L = 37.399041, acc = 0.870000\n",
      "L = 37.392325, acc = 0.870000\n",
      "L = 37.385568, acc = 0.865000\n",
      "L = 37.378772, acc = 0.865000\n",
      "L = 37.371935, acc = 0.865000\n",
      "L = 37.365057, acc = 0.865000\n",
      "L = 37.358138, acc = 0.865000\n",
      "L = 37.351177, acc = 0.865000\n",
      "L = 37.344175, acc = 0.865000\n",
      "L = 37.337130, acc = 0.865000\n",
      "L = 37.330043, acc = 0.865000\n",
      "L = 37.322913, acc = 0.865000\n",
      "L = 37.315740, acc = 0.865000\n",
      "L = 37.308524, acc = 0.865000\n",
      "L = 37.301263, acc = 0.865000\n",
      "L = 37.293959, acc = 0.865000\n",
      "L = 37.286609, acc = 0.865000\n",
      "L = 37.279215, acc = 0.865000\n",
      "L = 37.271776, acc = 0.865000\n",
      "L = 37.264291, acc = 0.865000\n",
      "L = 37.256760, acc = 0.865000\n",
      "L = 37.249183, acc = 0.865000\n",
      "L = 37.241559, acc = 0.865000\n",
      "L = 37.233887, acc = 0.865000\n",
      "L = 37.226168, acc = 0.865000\n",
      "L = 37.218402, acc = 0.865000\n",
      "L = 37.210587, acc = 0.865000\n",
      "L = 37.202723, acc = 0.865000\n",
      "L = 37.194810, acc = 0.865000\n",
      "L = 37.186848, acc = 0.865000\n",
      "L = 37.178835, acc = 0.865000\n",
      "L = 37.170773, acc = 0.865000\n",
      "L = 37.162659, acc = 0.865000\n",
      "L = 37.154495, acc = 0.865000\n",
      "L = 37.146279, acc = 0.865000\n",
      "L = 37.138011, acc = 0.865000\n",
      "L = 37.129691, acc = 0.865000\n",
      "L = 37.121318, acc = 0.865000\n",
      "L = 37.112892, acc = 0.865000\n",
      "L = 37.104412, acc = 0.865000\n",
      "L = 37.095878, acc = 0.865000\n",
      "L = 37.087289, acc = 0.865000\n",
      "L = 37.078646, acc = 0.865000\n",
      "L = 37.069947, acc = 0.865000\n",
      "L = 37.061192, acc = 0.865000\n",
      "L = 37.052381, acc = 0.865000\n",
      "L = 37.043513, acc = 0.865000\n",
      "L = 37.034588, acc = 0.865000\n",
      "L = 37.025606, acc = 0.865000\n",
      "L = 37.016565, acc = 0.865000\n",
      "L = 37.007466, acc = 0.865000\n",
      "L = 36.998308, acc = 0.865000\n",
      "L = 36.989090, acc = 0.865000\n",
      "L = 36.979813, acc = 0.865000\n",
      "L = 36.970475, acc = 0.865000\n",
      "L = 36.961076, acc = 0.865000\n",
      "L = 36.951616, acc = 0.865000\n",
      "L = 36.942094, acc = 0.865000\n",
      "L = 36.932510, acc = 0.865000\n",
      "L = 36.922864, acc = 0.865000\n",
      "L = 36.913154, acc = 0.865000\n",
      "L = 36.903380, acc = 0.865000\n",
      "L = 36.893542, acc = 0.865000\n",
      "L = 36.883640, acc = 0.865000\n",
      "L = 36.873672, acc = 0.865000\n",
      "L = 36.863639, acc = 0.865000\n",
      "L = 36.853540, acc = 0.865000\n",
      "L = 36.843375, acc = 0.865000\n",
      "L = 36.833142, acc = 0.865000\n",
      "L = 36.822842, acc = 0.865000\n",
      "L = 36.812474, acc = 0.865000\n",
      "L = 36.802038, acc = 0.865000\n",
      "L = 36.791533, acc = 0.865000\n",
      "L = 36.780958, acc = 0.865000\n",
      "L = 36.770314, acc = 0.865000\n",
      "L = 36.759599, acc = 0.865000\n",
      "L = 36.748813, acc = 0.865000\n",
      "L = 36.737956, acc = 0.865000\n",
      "L = 36.727028, acc = 0.865000\n",
      "L = 36.716027, acc = 0.865000\n",
      "L = 36.704953, acc = 0.865000\n",
      "L = 36.693806, acc = 0.865000\n",
      "L = 36.682585, acc = 0.865000\n",
      "L = 36.671291, acc = 0.865000\n",
      "L = 36.659921, acc = 0.865000\n",
      "L = 36.648477, acc = 0.865000\n",
      "L = 36.636957, acc = 0.865000\n",
      "L = 36.625361, acc = 0.865000\n",
      "L = 36.613689, acc = 0.865000\n",
      "L = 36.601939, acc = 0.865000\n",
      "L = 36.590112, acc = 0.865000\n",
      "L = 36.578208, acc = 0.865000\n",
      "L = 36.566225, acc = 0.865000\n",
      "L = 36.554163, acc = 0.865000\n",
      "L = 36.542022, acc = 0.870000\n",
      "L = 36.529801, acc = 0.870000\n",
      "L = 36.517501, acc = 0.870000\n",
      "L = 36.505119, acc = 0.870000\n",
      "L = 36.492657, acc = 0.870000\n",
      "L = 36.480113, acc = 0.870000\n",
      "L = 36.467487, acc = 0.870000\n",
      "L = 36.454780, acc = 0.870000\n",
      "L = 36.441989, acc = 0.870000\n",
      "L = 36.429115, acc = 0.870000\n",
      "L = 36.416158, acc = 0.870000\n",
      "L = 36.403116, acc = 0.870000\n",
      "L = 36.389990, acc = 0.870000\n",
      "L = 36.376780, acc = 0.870000\n",
      "L = 36.363484, acc = 0.870000\n",
      "L = 36.350103, acc = 0.870000\n",
      "L = 36.336635, acc = 0.870000\n",
      "L = 36.323081, acc = 0.870000\n",
      "L = 36.309441, acc = 0.870000\n",
      "L = 36.295713, acc = 0.870000\n",
      "L = 36.281898, acc = 0.870000\n",
      "L = 36.267995, acc = 0.870000\n",
      "L = 36.254003, acc = 0.870000\n",
      "L = 36.239923, acc = 0.870000\n",
      "L = 36.225754, acc = 0.870000\n",
      "L = 36.211496, acc = 0.870000\n",
      "L = 36.197148, acc = 0.870000\n",
      "L = 36.182710, acc = 0.870000\n",
      "L = 36.168182, acc = 0.870000\n",
      "L = 36.153563, acc = 0.870000\n",
      "L = 36.138853, acc = 0.870000\n",
      "L = 36.124052, acc = 0.870000\n",
      "L = 36.109160, acc = 0.870000\n",
      "L = 36.094175, acc = 0.870000\n",
      "L = 36.079099, acc = 0.870000\n",
      "L = 36.063930, acc = 0.870000\n",
      "L = 36.048668, acc = 0.870000\n",
      "L = 36.033313, acc = 0.870000\n",
      "L = 36.017865, acc = 0.870000\n",
      "L = 36.002323, acc = 0.870000\n",
      "L = 35.986688, acc = 0.870000\n",
      "L = 35.970959, acc = 0.870000\n",
      "L = 35.955135, acc = 0.870000\n",
      "L = 35.939218, acc = 0.870000\n",
      "L = 35.923205, acc = 0.870000\n",
      "L = 35.907098, acc = 0.870000\n",
      "L = 35.890895, acc = 0.870000\n",
      "L = 35.874597, acc = 0.870000\n",
      "L = 35.858204, acc = 0.870000\n",
      "L = 35.841715, acc = 0.870000\n",
      "L = 35.825131, acc = 0.875000\n",
      "L = 35.808450, acc = 0.875000\n",
      "L = 35.791674, acc = 0.875000\n",
      "L = 35.774801, acc = 0.875000\n",
      "L = 35.757831, acc = 0.875000\n",
      "L = 35.740766, acc = 0.875000\n",
      "L = 35.723603, acc = 0.875000\n",
      "L = 35.706344, acc = 0.875000\n",
      "L = 35.688988, acc = 0.875000\n",
      "L = 35.671535, acc = 0.875000\n",
      "L = 35.653986, acc = 0.875000\n",
      "L = 35.636339, acc = 0.875000\n",
      "L = 35.618595, acc = 0.875000\n",
      "L = 35.600754, acc = 0.875000\n",
      "L = 35.582815, acc = 0.875000\n",
      "L = 35.564779, acc = 0.875000\n",
      "L = 35.546646, acc = 0.875000\n",
      "L = 35.528416, acc = 0.875000\n",
      "L = 35.510088, acc = 0.875000\n",
      "L = 35.491663, acc = 0.875000\n",
      "L = 35.473141, acc = 0.875000\n",
      "L = 35.454522, acc = 0.875000\n",
      "L = 35.435805, acc = 0.880000\n",
      "L = 35.416991, acc = 0.880000\n",
      "L = 35.398079, acc = 0.880000\n",
      "L = 35.379071, acc = 0.880000\n",
      "L = 35.359966, acc = 0.880000\n",
      "L = 35.340763, acc = 0.880000\n",
      "L = 35.321464, acc = 0.880000\n",
      "L = 35.302068, acc = 0.880000\n",
      "L = 35.282575, acc = 0.880000\n",
      "L = 35.262986, acc = 0.880000\n",
      "L = 35.243300, acc = 0.880000\n",
      "L = 35.223518, acc = 0.880000\n",
      "L = 35.203640, acc = 0.880000\n",
      "L = 35.183666, acc = 0.880000\n",
      "L = 35.163596, acc = 0.880000\n",
      "L = 35.143430, acc = 0.880000\n",
      "L = 35.123168, acc = 0.880000\n",
      "L = 35.102812, acc = 0.880000\n",
      "L = 35.082360, acc = 0.880000\n",
      "L = 35.061813, acc = 0.880000\n",
      "L = 35.041172, acc = 0.880000\n",
      "L = 35.020436, acc = 0.880000\n",
      "L = 34.999606, acc = 0.880000\n",
      "L = 34.978682, acc = 0.885000\n",
      "L = 34.957664, acc = 0.885000\n",
      "L = 34.936552, acc = 0.885000\n",
      "L = 34.915348, acc = 0.885000\n",
      "L = 34.894050, acc = 0.885000\n",
      "L = 34.872660, acc = 0.885000\n",
      "L = 34.851178, acc = 0.885000\n",
      "L = 34.829603, acc = 0.885000\n",
      "L = 34.807937, acc = 0.885000\n",
      "L = 34.786179, acc = 0.885000\n",
      "L = 34.764330, acc = 0.885000\n",
      "L = 34.742391, acc = 0.885000\n",
      "L = 34.720361, acc = 0.885000\n",
      "L = 34.698241, acc = 0.885000\n",
      "L = 34.676031, acc = 0.885000\n",
      "L = 34.653732, acc = 0.885000\n",
      "L = 34.631344, acc = 0.885000\n",
      "L = 34.608868, acc = 0.885000\n",
      "L = 34.586304, acc = 0.885000\n",
      "L = 34.563651, acc = 0.885000\n",
      "L = 34.540912, acc = 0.885000\n",
      "L = 34.518085, acc = 0.885000\n",
      "L = 34.495172, acc = 0.885000\n",
      "L = 34.472173, acc = 0.885000\n",
      "L = 34.449089, acc = 0.885000\n",
      "L = 34.425919, acc = 0.885000\n",
      "L = 34.402665, acc = 0.885000\n",
      "L = 34.379326, acc = 0.885000\n",
      "L = 34.355904, acc = 0.885000\n",
      "L = 34.332398, acc = 0.885000\n",
      "L = 34.308810, acc = 0.885000\n",
      "L = 34.285139, acc = 0.885000\n",
      "L = 34.261386, acc = 0.885000\n",
      "L = 34.237552, acc = 0.885000\n",
      "L = 34.213637, acc = 0.885000\n",
      "L = 34.189642, acc = 0.885000\n",
      "L = 34.165567, acc = 0.885000\n",
      "L = 34.141413, acc = 0.885000\n",
      "L = 34.117180, acc = 0.885000\n",
      "L = 34.092869, acc = 0.885000\n",
      "L = 34.068481, acc = 0.885000\n",
      "L = 34.044015, acc = 0.885000\n",
      "L = 34.019472, acc = 0.885000\n",
      "L = 33.994854, acc = 0.885000\n",
      "L = 33.970160, acc = 0.885000\n",
      "L = 33.945392, acc = 0.890000\n",
      "L = 33.920549, acc = 0.890000\n",
      "L = 33.895632, acc = 0.890000\n",
      "L = 33.870642, acc = 0.890000\n",
      "L = 33.845580, acc = 0.890000\n",
      "L = 33.820446, acc = 0.890000\n",
      "L = 33.795240, acc = 0.890000\n",
      "L = 33.769964, acc = 0.890000\n",
      "L = 33.744617, acc = 0.890000\n",
      "L = 33.719201, acc = 0.890000\n",
      "L = 33.693716, acc = 0.890000\n",
      "L = 33.668162, acc = 0.890000\n",
      "L = 33.642541, acc = 0.890000\n",
      "L = 33.616852, acc = 0.890000\n",
      "L = 33.591098, acc = 0.890000\n",
      "L = 33.565277, acc = 0.890000\n",
      "L = 33.539391, acc = 0.890000\n",
      "L = 33.513440, acc = 0.890000\n",
      "L = 33.487426, acc = 0.890000\n",
      "L = 33.461348, acc = 0.890000\n",
      "L = 33.435207, acc = 0.890000\n",
      "L = 33.409004, acc = 0.890000\n",
      "L = 33.382740, acc = 0.890000\n",
      "L = 33.356415, acc = 0.890000\n",
      "L = 33.330030, acc = 0.890000\n",
      "L = 33.303586, acc = 0.890000\n",
      "L = 33.277082, acc = 0.890000\n",
      "L = 33.250521, acc = 0.890000\n",
      "L = 33.223901, acc = 0.890000\n",
      "L = 33.197225, acc = 0.890000\n",
      "L = 33.170493, acc = 0.890000\n",
      "L = 33.143705, acc = 0.890000\n",
      "L = 33.116862, acc = 0.890000\n",
      "L = 33.089964, acc = 0.890000\n",
      "L = 33.063013, acc = 0.890000\n",
      "L = 33.036009, acc = 0.890000\n",
      "L = 33.008953, acc = 0.890000\n",
      "L = 32.981845, acc = 0.890000\n",
      "L = 32.954686, acc = 0.890000\n",
      "L = 32.927477, acc = 0.890000\n",
      "L = 32.900218, acc = 0.890000\n",
      "L = 32.872910, acc = 0.890000\n",
      "L = 32.845553, acc = 0.890000\n",
      "L = 32.818149, acc = 0.890000\n",
      "L = 32.790697, acc = 0.890000\n",
      "L = 32.763199, acc = 0.890000\n",
      "L = 32.735656, acc = 0.890000\n",
      "L = 32.708067, acc = 0.890000\n",
      "L = 32.680434, acc = 0.890000\n",
      "L = 32.652757, acc = 0.890000\n",
      "L = 32.625036, acc = 0.890000\n",
      "L = 32.597273, acc = 0.890000\n",
      "L = 32.569468, acc = 0.890000\n",
      "L = 32.541622, acc = 0.890000\n",
      "L = 32.513736, acc = 0.890000\n",
      "L = 32.485809, acc = 0.890000\n",
      "L = 32.457843, acc = 0.890000\n",
      "L = 32.429838, acc = 0.890000\n",
      "L = 32.401796, acc = 0.890000\n",
      "L = 32.373715, acc = 0.890000\n",
      "L = 32.345598, acc = 0.890000\n",
      "L = 32.317445, acc = 0.890000\n",
      "L = 32.289257, acc = 0.890000\n",
      "L = 32.261033, acc = 0.890000\n",
      "L = 32.232775, acc = 0.890000\n",
      "L = 32.204484, acc = 0.890000\n",
      "L = 32.176159, acc = 0.890000\n",
      "L = 32.147802, acc = 0.890000\n",
      "L = 32.119413, acc = 0.890000\n",
      "L = 32.090993, acc = 0.890000\n",
      "L = 32.062542, acc = 0.890000\n",
      "L = 32.034062, acc = 0.890000\n",
      "L = 32.005552, acc = 0.890000\n",
      "L = 31.977013, acc = 0.890000\n",
      "L = 31.948446, acc = 0.890000\n",
      "L = 31.919852, acc = 0.890000\n",
      "L = 31.891230, acc = 0.890000\n",
      "L = 31.862582, acc = 0.890000\n",
      "L = 31.833908, acc = 0.890000\n",
      "L = 31.805209, acc = 0.890000\n",
      "L = 31.776486, acc = 0.895000\n",
      "L = 31.747738, acc = 0.895000\n",
      "L = 31.718967, acc = 0.895000\n",
      "L = 31.690173, acc = 0.895000\n",
      "L = 31.661356, acc = 0.895000\n",
      "L = 31.632518, acc = 0.895000\n",
      "L = 31.603659, acc = 0.895000\n",
      "L = 31.574779, acc = 0.895000\n",
      "L = 31.545878, acc = 0.895000\n",
      "L = 31.516959, acc = 0.895000\n",
      "L = 31.488020, acc = 0.895000\n",
      "L = 31.459062, acc = 0.895000\n",
      "L = 31.430087, acc = 0.895000\n",
      "L = 31.401095, acc = 0.895000\n",
      "L = 31.372085, acc = 0.895000\n",
      "L = 31.343059, acc = 0.895000\n",
      "L = 31.314018, acc = 0.900000\n",
      "L = 31.284961, acc = 0.900000\n",
      "L = 31.255889, acc = 0.900000\n",
      "L = 31.226803, acc = 0.900000\n",
      "L = 31.197704, acc = 0.900000\n",
      "L = 31.168591, acc = 0.900000\n",
      "L = 31.139466, acc = 0.900000\n",
      "L = 31.110328, acc = 0.900000\n",
      "L = 31.081178, acc = 0.900000\n",
      "L = 31.052018, acc = 0.900000\n",
      "L = 31.022847, acc = 0.900000\n",
      "L = 30.993665, acc = 0.900000\n",
      "L = 30.964474, acc = 0.900000\n",
      "L = 30.935273, acc = 0.900000\n",
      "L = 30.906064, acc = 0.900000\n",
      "L = 30.876846, acc = 0.900000\n",
      "L = 30.847620, acc = 0.900000\n",
      "L = 30.818387, acc = 0.900000\n",
      "L = 30.789148, acc = 0.900000\n",
      "L = 30.759901, acc = 0.900000\n",
      "L = 30.730649, acc = 0.900000\n",
      "L = 30.701391, acc = 0.900000\n",
      "L = 30.672128, acc = 0.900000\n",
      "L = 30.642860, acc = 0.900000\n",
      "L = 30.613588, acc = 0.900000\n",
      "L = 30.584312, acc = 0.900000\n",
      "L = 30.555033, acc = 0.905000\n",
      "L = 30.525751, acc = 0.905000\n",
      "L = 30.496466, acc = 0.905000\n",
      "L = 30.467180, acc = 0.905000\n",
      "L = 30.437891, acc = 0.905000\n",
      "L = 30.408601, acc = 0.905000\n",
      "L = 30.379311, acc = 0.905000\n",
      "L = 30.350020, acc = 0.905000\n",
      "L = 30.320729, acc = 0.905000\n",
      "L = 30.291438, acc = 0.905000\n",
      "L = 30.262148, acc = 0.905000\n",
      "L = 30.232859, acc = 0.905000\n",
      "L = 30.203571, acc = 0.905000\n",
      "L = 30.174285, acc = 0.905000\n",
      "L = 30.145002, acc = 0.905000\n",
      "L = 30.115721, acc = 0.910000\n",
      "L = 30.086443, acc = 0.910000\n",
      "L = 30.057169, acc = 0.910000\n",
      "L = 30.027898, acc = 0.910000\n",
      "L = 29.998632, acc = 0.910000\n",
      "L = 29.969370, acc = 0.910000\n",
      "L = 29.940112, acc = 0.910000\n",
      "L = 29.910860, acc = 0.910000\n",
      "L = 29.881613, acc = 0.910000\n",
      "L = 29.852372, acc = 0.910000\n",
      "L = 29.823137, acc = 0.910000\n",
      "L = 29.793909, acc = 0.910000\n",
      "L = 29.764688, acc = 0.910000\n",
      "L = 29.735473, acc = 0.910000\n",
      "L = 29.706266, acc = 0.910000\n",
      "L = 29.677067, acc = 0.910000\n",
      "L = 29.647876, acc = 0.910000\n",
      "L = 29.618694, acc = 0.910000\n",
      "L = 29.589520, acc = 0.910000\n",
      "L = 29.560355, acc = 0.910000\n",
      "L = 29.531200, acc = 0.910000\n",
      "L = 29.502054, acc = 0.910000\n",
      "L = 29.472918, acc = 0.910000\n",
      "L = 29.443792, acc = 0.910000\n",
      "L = 29.414677, acc = 0.910000\n",
      "L = 29.385573, acc = 0.910000\n",
      "L = 29.356479, acc = 0.910000\n",
      "L = 29.327397, acc = 0.910000\n",
      "L = 29.298327, acc = 0.910000\n",
      "L = 29.269269, acc = 0.910000\n",
      "L = 29.240222, acc = 0.910000\n",
      "L = 29.211189, acc = 0.910000\n",
      "L = 29.182167, acc = 0.915000\n",
      "L = 29.153159, acc = 0.915000\n",
      "L = 29.124164, acc = 0.915000\n",
      "L = 29.095183, acc = 0.915000\n",
      "L = 29.066215, acc = 0.915000\n",
      "L = 29.037261, acc = 0.915000\n",
      "L = 29.008321, acc = 0.915000\n",
      "L = 28.979396, acc = 0.915000\n",
      "L = 28.950485, acc = 0.915000\n",
      "L = 28.921590, acc = 0.915000\n",
      "L = 28.892709, acc = 0.915000\n",
      "L = 28.863844, acc = 0.915000\n",
      "L = 28.834994, acc = 0.915000\n",
      "L = 28.806161, acc = 0.915000\n",
      "L = 28.777343, acc = 0.915000\n",
      "L = 28.748541, acc = 0.915000\n",
      "L = 28.719756, acc = 0.915000\n",
      "L = 28.690988, acc = 0.915000\n",
      "L = 28.662236, acc = 0.915000\n",
      "L = 28.633502, acc = 0.915000\n",
      "L = 28.604784, acc = 0.915000\n",
      "L = 28.576085, acc = 0.915000\n",
      "L = 28.547403, acc = 0.915000\n",
      "L = 28.518738, acc = 0.915000\n",
      "L = 28.490092, acc = 0.915000\n",
      "L = 28.461464, acc = 0.915000\n",
      "L = 28.432854, acc = 0.915000\n",
      "L = 28.404263, acc = 0.915000\n",
      "L = 28.375691, acc = 0.915000\n",
      "L = 28.347138, acc = 0.915000\n",
      "L = 28.318604, acc = 0.915000\n",
      "L = 28.290089, acc = 0.915000\n",
      "L = 28.261594, acc = 0.915000\n",
      "L = 28.233118, acc = 0.915000\n",
      "L = 28.204662, acc = 0.915000\n",
      "L = 28.176226, acc = 0.915000\n",
      "L = 28.147810, acc = 0.915000\n",
      "L = 28.119415, acc = 0.915000\n",
      "L = 28.091039, acc = 0.915000\n",
      "L = 28.062685, acc = 0.915000\n",
      "L = 28.034351, acc = 0.915000\n",
      "L = 28.006038, acc = 0.915000\n",
      "L = 27.977746, acc = 0.915000\n",
      "L = 27.949475, acc = 0.915000\n",
      "L = 27.921225, acc = 0.915000\n",
      "L = 27.892997, acc = 0.915000\n",
      "L = 27.864791, acc = 0.915000\n",
      "L = 27.836606, acc = 0.915000\n",
      "L = 27.808443, acc = 0.915000\n",
      "L = 27.780302, acc = 0.915000\n",
      "L = 27.752183, acc = 0.915000\n",
      "L = 27.724086, acc = 0.915000\n",
      "L = 27.696011, acc = 0.915000\n",
      "L = 27.667960, acc = 0.915000\n",
      "L = 27.639930, acc = 0.915000\n",
      "L = 27.611923, acc = 0.915000\n",
      "L = 27.583939, acc = 0.915000\n",
      "L = 27.555978, acc = 0.915000\n",
      "L = 27.528040, acc = 0.915000\n",
      "L = 27.500125, acc = 0.915000\n",
      "L = 27.472234, acc = 0.915000\n",
      "L = 27.444365, acc = 0.915000\n",
      "L = 27.416521, acc = 0.915000\n",
      "L = 27.388699, acc = 0.915000\n",
      "L = 27.360902, acc = 0.915000\n",
      "L = 27.333128, acc = 0.915000\n",
      "L = 27.305378, acc = 0.915000\n",
      "L = 27.277651, acc = 0.915000\n",
      "L = 27.249949, acc = 0.915000\n",
      "L = 27.222271, acc = 0.915000\n",
      "L = 27.194617, acc = 0.915000\n",
      "L = 27.166988, acc = 0.915000\n",
      "L = 27.139383, acc = 0.915000\n",
      "L = 27.111802, acc = 0.915000\n",
      "L = 27.084246, acc = 0.915000\n",
      "L = 27.056714, acc = 0.915000\n",
      "L = 27.029207, acc = 0.915000\n",
      "L = 27.001725, acc = 0.915000\n",
      "L = 26.974268, acc = 0.915000\n",
      "L = 26.946836, acc = 0.915000\n",
      "L = 26.919428, acc = 0.915000\n",
      "L = 26.892046, acc = 0.915000\n",
      "L = 26.864689, acc = 0.915000\n",
      "L = 26.837357, acc = 0.920000\n",
      "L = 26.810050, acc = 0.920000\n",
      "L = 26.782769, acc = 0.920000\n",
      "L = 26.755513, acc = 0.920000\n",
      "L = 26.728283, acc = 0.920000\n",
      "L = 26.701078, acc = 0.920000\n",
      "L = 26.673898, acc = 0.920000\n",
      "L = 26.646745, acc = 0.920000\n",
      "L = 26.619617, acc = 0.920000\n",
      "L = 26.592515, acc = 0.920000\n",
      "L = 26.565438, acc = 0.920000\n",
      "L = 26.538388, acc = 0.925000\n",
      "L = 26.511363, acc = 0.925000\n",
      "L = 26.484365, acc = 0.925000\n",
      "L = 26.457392, acc = 0.925000\n",
      "L = 26.430446, acc = 0.925000\n",
      "L = 26.403525, acc = 0.925000\n",
      "L = 26.376631, acc = 0.925000\n",
      "L = 26.349763, acc = 0.925000\n",
      "L = 26.322922, acc = 0.925000\n",
      "L = 26.296106, acc = 0.925000\n",
      "L = 26.269317, acc = 0.925000\n",
      "L = 26.242555, acc = 0.925000\n",
      "L = 26.215819, acc = 0.925000\n",
      "L = 26.189109, acc = 0.925000\n",
      "L = 26.162426, acc = 0.925000\n",
      "L = 26.135770, acc = 0.925000\n",
      "L = 26.109140, acc = 0.925000\n",
      "L = 26.082537, acc = 0.925000\n",
      "L = 26.055960, acc = 0.925000\n",
      "L = 26.029411, acc = 0.925000\n",
      "L = 26.002888, acc = 0.925000\n",
      "L = 25.976392, acc = 0.925000\n",
      "L = 25.949922, acc = 0.925000\n",
      "L = 25.923480, acc = 0.925000\n",
      "L = 25.897064, acc = 0.925000\n",
      "L = 25.870676, acc = 0.925000\n",
      "L = 25.844314, acc = 0.925000\n",
      "L = 25.817979, acc = 0.925000\n",
      "L = 25.791672, acc = 0.925000\n",
      "L = 25.765391, acc = 0.925000\n",
      "L = 25.739137, acc = 0.925000\n",
      "L = 25.712911, acc = 0.930000\n",
      "L = 25.686712, acc = 0.935000\n",
      "L = 25.660540, acc = 0.935000\n",
      "L = 25.634395, acc = 0.935000\n",
      "L = 25.608277, acc = 0.935000\n",
      "L = 25.582187, acc = 0.935000\n",
      "L = 25.556124, acc = 0.935000\n",
      "L = 25.530088, acc = 0.935000\n",
      "L = 25.504079, acc = 0.935000\n",
      "L = 25.478098, acc = 0.935000\n",
      "L = 25.452144, acc = 0.935000\n",
      "L = 25.426218, acc = 0.935000\n",
      "L = 25.400319, acc = 0.935000\n",
      "L = 25.374448, acc = 0.935000\n",
      "L = 25.348604, acc = 0.935000\n",
      "L = 25.322787, acc = 0.935000\n",
      "L = 25.296998, acc = 0.935000\n",
      "L = 25.271237, acc = 0.935000\n",
      "L = 25.245503, acc = 0.935000\n",
      "L = 25.219796, acc = 0.935000\n",
      "L = 25.194118, acc = 0.935000\n",
      "L = 25.168466, acc = 0.935000\n",
      "L = 25.142843, acc = 0.935000\n",
      "L = 25.117247, acc = 0.935000\n",
      "L = 25.091679, acc = 0.935000\n",
      "L = 25.066138, acc = 0.935000\n",
      "L = 25.040626, acc = 0.935000\n",
      "L = 25.015141, acc = 0.935000\n",
      "L = 24.989683, acc = 0.935000\n",
      "L = 24.964254, acc = 0.935000\n",
      "L = 24.938852, acc = 0.935000\n",
      "L = 24.913478, acc = 0.935000\n",
      "L = 24.888132, acc = 0.935000\n",
      "L = 24.862814, acc = 0.935000\n",
      "L = 24.837523, acc = 0.935000\n",
      "L = 24.812261, acc = 0.935000\n",
      "L = 24.787026, acc = 0.935000\n",
      "L = 24.761820, acc = 0.935000\n",
      "L = 24.736641, acc = 0.935000\n",
      "L = 24.711490, acc = 0.935000\n",
      "L = 24.686367, acc = 0.935000\n",
      "L = 24.661272, acc = 0.935000\n",
      "L = 24.636206, acc = 0.935000\n",
      "L = 24.611167, acc = 0.935000\n",
      "L = 24.586156, acc = 0.935000\n",
      "L = 24.561173, acc = 0.935000\n",
      "L = 24.536219, acc = 0.935000\n",
      "L = 24.511292, acc = 0.935000\n",
      "L = 24.486394, acc = 0.935000\n",
      "L = 24.461523, acc = 0.935000\n",
      "L = 24.436681, acc = 0.935000\n",
      "L = 24.411867, acc = 0.935000\n",
      "L = 24.387081, acc = 0.935000\n",
      "L = 24.362324, acc = 0.935000\n",
      "L = 24.337594, acc = 0.935000\n",
      "L = 24.312893, acc = 0.935000\n",
      "L = 24.288220, acc = 0.935000\n",
      "L = 24.263576, acc = 0.935000\n",
      "L = 24.238959, acc = 0.935000\n",
      "L = 24.214371, acc = 0.935000\n",
      "L = 24.189812, acc = 0.935000\n",
      "L = 24.165280, acc = 0.935000\n",
      "L = 24.140777, acc = 0.935000\n",
      "L = 24.116303, acc = 0.935000\n",
      "L = 24.091857, acc = 0.935000\n",
      "L = 24.067439, acc = 0.940000\n",
      "L = 24.043050, acc = 0.940000\n",
      "L = 24.018689, acc = 0.940000\n",
      "L = 23.994357, acc = 0.940000\n",
      "L = 23.970053, acc = 0.940000\n",
      "L = 23.945777, acc = 0.940000\n",
      "L = 23.921531, acc = 0.940000\n",
      "L = 23.897312, acc = 0.940000\n",
      "L = 23.873123, acc = 0.940000\n",
      "L = 23.848962, acc = 0.940000\n",
      "L = 23.824830, acc = 0.940000\n",
      "L = 23.800726, acc = 0.940000\n",
      "L = 23.776651, acc = 0.940000\n",
      "L = 23.752605, acc = 0.940000\n",
      "L = 23.728587, acc = 0.940000\n",
      "L = 23.704598, acc = 0.940000\n",
      "L = 23.680638, acc = 0.940000\n",
      "L = 23.656707, acc = 0.940000\n",
      "L = 23.632804, acc = 0.940000\n",
      "L = 23.608931, acc = 0.940000\n",
      "L = 23.585086, acc = 0.940000\n",
      "L = 23.561270, acc = 0.940000\n",
      "L = 23.537483, acc = 0.940000\n",
      "L = 23.513725, acc = 0.940000\n",
      "L = 23.489996, acc = 0.940000\n",
      "L = 23.466296, acc = 0.940000\n",
      "L = 23.442625, acc = 0.940000\n",
      "L = 23.418983, acc = 0.940000\n",
      "L = 23.395370, acc = 0.940000\n",
      "L = 23.371786, acc = 0.940000\n",
      "L = 23.348231, acc = 0.940000\n",
      "L = 23.324705, acc = 0.940000\n",
      "L = 23.301209, acc = 0.940000\n",
      "L = 23.277741, acc = 0.940000\n",
      "L = 23.254303, acc = 0.940000\n",
      "L = 23.230894, acc = 0.940000\n",
      "L = 23.207515, acc = 0.940000\n",
      "L = 23.184164, acc = 0.940000\n",
      "L = 23.160843, acc = 0.940000\n",
      "L = 23.137552, acc = 0.940000\n",
      "L = 23.114289, acc = 0.940000\n",
      "L = 23.091057, acc = 0.940000\n",
      "L = 23.067853, acc = 0.940000\n",
      "L = 23.044679, acc = 0.940000\n",
      "L = 23.021535, acc = 0.940000\n",
      "L = 22.998420, acc = 0.940000\n",
      "L = 22.975335, acc = 0.940000\n",
      "L = 22.952279, acc = 0.940000\n",
      "L = 22.929253, acc = 0.940000\n",
      "L = 22.906256, acc = 0.940000\n",
      "L = 22.883289, acc = 0.940000\n",
      "L = 22.860352, acc = 0.940000\n",
      "L = 22.837445, acc = 0.940000\n",
      "L = 22.814567, acc = 0.940000\n",
      "L = 22.791719, acc = 0.940000\n",
      "L = 22.768901, acc = 0.940000\n",
      "L = 22.746113, acc = 0.940000\n",
      "L = 22.723355, acc = 0.940000\n",
      "L = 22.700627, acc = 0.940000\n",
      "L = 22.677928, acc = 0.940000\n",
      "L = 22.655260, acc = 0.940000\n",
      "L = 22.632622, acc = 0.940000\n",
      "L = 22.610014, acc = 0.940000\n",
      "L = 22.587435, acc = 0.940000\n",
      "L = 22.564887, acc = 0.940000\n",
      "L = 22.542369, acc = 0.940000\n",
      "L = 22.519882, acc = 0.940000\n",
      "L = 22.497424, acc = 0.940000\n",
      "L = 22.474997, acc = 0.940000\n",
      "L = 22.452600, acc = 0.940000\n",
      "L = 22.430234, acc = 0.940000\n",
      "L = 22.407897, acc = 0.940000\n",
      "L = 22.385592, acc = 0.940000\n",
      "L = 22.363316, acc = 0.940000\n",
      "L = 22.341071, acc = 0.940000\n",
      "L = 22.318857, acc = 0.940000\n",
      "L = 22.296673, acc = 0.940000\n",
      "L = 22.274519, acc = 0.940000\n",
      "L = 22.252397, acc = 0.940000\n",
      "L = 22.230305, acc = 0.940000\n",
      "L = 22.208243, acc = 0.940000\n",
      "L = 22.186212, acc = 0.940000\n",
      "L = 22.164212, acc = 0.940000\n",
      "L = 22.142243, acc = 0.940000\n",
      "L = 22.120305, acc = 0.940000\n",
      "L = 22.098397, acc = 0.940000\n",
      "L = 22.076521, acc = 0.945000\n",
      "L = 22.054675, acc = 0.945000\n",
      "L = 22.032860, acc = 0.945000\n",
      "L = 22.011076, acc = 0.945000\n",
      "L = 21.989323, acc = 0.945000\n",
      "L = 21.967602, acc = 0.945000\n",
      "L = 21.945911, acc = 0.945000\n",
      "L = 21.924252, acc = 0.945000\n",
      "L = 21.902623, acc = 0.945000\n",
      "L = 21.881026, acc = 0.945000\n",
      "L = 21.859460, acc = 0.945000\n",
      "L = 21.837925, acc = 0.945000\n",
      "L = 21.816422, acc = 0.945000\n",
      "L = 21.794950, acc = 0.945000\n",
      "L = 21.773509, acc = 0.945000\n",
      "L = 21.752100, acc = 0.945000\n",
      "L = 21.730722, acc = 0.945000\n",
      "L = 21.709376, acc = 0.945000\n",
      "L = 21.688061, acc = 0.945000\n",
      "L = 21.666778, acc = 0.945000\n",
      "L = 21.645526, acc = 0.945000\n",
      "L = 21.624306, acc = 0.945000\n",
      "L = 21.603118, acc = 0.945000\n",
      "L = 21.581961, acc = 0.945000\n",
      "L = 21.560836, acc = 0.945000\n",
      "L = 21.539743, acc = 0.945000\n",
      "L = 21.518682, acc = 0.945000\n",
      "L = 21.497652, acc = 0.945000\n",
      "L = 21.476654, acc = 0.945000\n",
      "L = 21.455688, acc = 0.945000\n",
      "L = 21.434754, acc = 0.945000\n",
      "L = 21.413852, acc = 0.945000\n",
      "L = 21.392982, acc = 0.945000\n",
      "L = 21.372144, acc = 0.945000\n",
      "L = 21.351339, acc = 0.945000\n",
      "L = 21.330565, acc = 0.945000\n",
      "L = 21.309823, acc = 0.945000\n",
      "L = 21.289113, acc = 0.945000\n",
      "L = 21.268436, acc = 0.945000\n",
      "L = 21.247791, acc = 0.945000\n",
      "L = 21.227178, acc = 0.945000\n",
      "L = 21.206597, acc = 0.945000\n",
      "L = 21.186049, acc = 0.945000\n",
      "L = 21.165533, acc = 0.945000\n",
      "L = 21.145049, acc = 0.945000\n",
      "L = 21.124598, acc = 0.945000\n",
      "L = 21.104179, acc = 0.945000\n",
      "L = 21.083793, acc = 0.945000\n",
      "L = 21.063439, acc = 0.945000\n",
      "L = 21.043118, acc = 0.945000\n",
      "L = 21.022829, acc = 0.945000\n",
      "L = 21.002573, acc = 0.945000\n",
      "L = 20.982349, acc = 0.945000\n",
      "L = 20.962158, acc = 0.945000\n",
      "L = 20.942000, acc = 0.945000\n",
      "L = 20.921874, acc = 0.945000\n",
      "L = 20.901781, acc = 0.945000\n",
      "L = 20.881721, acc = 0.945000\n",
      "L = 20.861693, acc = 0.945000\n",
      "L = 20.841699, acc = 0.945000\n",
      "L = 20.821737, acc = 0.945000\n",
      "L = 20.801808, acc = 0.945000\n",
      "L = 20.781912, acc = 0.945000\n",
      "L = 20.762049, acc = 0.945000\n",
      "L = 20.742218, acc = 0.945000\n",
      "L = 20.722421, acc = 0.945000\n",
      "L = 20.702656, acc = 0.945000\n",
      "L = 20.682925, acc = 0.945000\n",
      "L = 20.663227, acc = 0.945000\n",
      "L = 20.643561, acc = 0.945000\n",
      "L = 20.623929, acc = 0.945000\n",
      "L = 20.604330, acc = 0.945000\n",
      "L = 20.584763, acc = 0.945000\n",
      "L = 20.565230, acc = 0.945000\n",
      "L = 20.545730, acc = 0.945000\n",
      "L = 20.526264, acc = 0.945000\n",
      "L = 20.506830, acc = 0.945000\n",
      "L = 20.487430, acc = 0.945000\n",
      "L = 20.468063, acc = 0.945000\n",
      "L = 20.448729, acc = 0.945000\n",
      "L = 20.429428, acc = 0.945000\n",
      "L = 20.410160, acc = 0.945000\n",
      "L = 20.390926, acc = 0.945000\n",
      "L = 20.371726, acc = 0.945000\n",
      "L = 20.352558, acc = 0.945000\n",
      "L = 20.333424, acc = 0.945000\n",
      "L = 20.314323, acc = 0.945000\n",
      "L = 20.295256, acc = 0.945000\n",
      "L = 20.276221, acc = 0.945000\n",
      "L = 20.257221, acc = 0.945000\n",
      "L = 20.238253, acc = 0.945000\n",
      "L = 20.219320, acc = 0.945000\n",
      "L = 20.200419, acc = 0.945000\n",
      "L = 20.181552, acc = 0.945000\n",
      "L = 20.162719, acc = 0.945000\n",
      "L = 20.143918, acc = 0.945000\n",
      "L = 20.125152, acc = 0.945000\n",
      "L = 20.106419, acc = 0.945000\n",
      "L = 20.087719, acc = 0.945000\n",
      "L = 20.069053, acc = 0.945000\n",
      "L = 20.050420, acc = 0.945000\n",
      "L = 20.031821, acc = 0.945000\n",
      "L = 20.013256, acc = 0.945000\n",
      "L = 19.994724, acc = 0.945000\n",
      "L = 19.976225, acc = 0.945000\n",
      "L = 19.957760, acc = 0.945000\n",
      "L = 19.939329, acc = 0.945000\n",
      "L = 19.920931, acc = 0.945000\n",
      "L = 19.902567, acc = 0.945000\n",
      "L = 19.884236, acc = 0.945000\n",
      "L = 19.865939, acc = 0.945000\n",
      "L = 19.847675, acc = 0.945000\n",
      "L = 19.829445, acc = 0.945000\n",
      "L = 19.811249, acc = 0.945000\n",
      "L = 19.793086, acc = 0.945000\n",
      "L = 19.774957, acc = 0.945000\n",
      "L = 19.756861, acc = 0.945000\n",
      "L = 19.738799, acc = 0.945000\n",
      "L = 19.720771, acc = 0.945000\n",
      "L = 19.702776, acc = 0.945000\n",
      "L = 19.684815, acc = 0.945000\n",
      "L = 19.666887, acc = 0.945000\n",
      "L = 19.648993, acc = 0.945000\n",
      "L = 19.631132, acc = 0.945000\n",
      "L = 19.613305, acc = 0.945000\n",
      "L = 19.595512, acc = 0.945000\n",
      "L = 19.577752, acc = 0.945000\n",
      "L = 19.560026, acc = 0.945000\n",
      "L = 19.542333, acc = 0.945000\n",
      "L = 19.524674, acc = 0.945000\n",
      "L = 19.507048, acc = 0.945000\n",
      "L = 19.489456, acc = 0.945000\n",
      "L = 19.471897, acc = 0.945000\n",
      "L = 19.454372, acc = 0.945000\n",
      "L = 19.436881, acc = 0.945000\n",
      "L = 19.419423, acc = 0.945000\n",
      "L = 19.401998, acc = 0.945000\n",
      "L = 19.384607, acc = 0.945000\n",
      "L = 19.367249, acc = 0.945000\n",
      "L = 19.349925, acc = 0.945000\n",
      "L = 19.332634, acc = 0.945000\n",
      "L = 19.315377, acc = 0.945000\n",
      "L = 19.298153, acc = 0.945000\n",
      "L = 19.280963, acc = 0.945000\n",
      "L = 19.263806, acc = 0.945000\n",
      "L = 19.246682, acc = 0.945000\n",
      "L = 19.229592, acc = 0.945000\n",
      "L = 19.212535, acc = 0.945000\n",
      "L = 19.195511, acc = 0.945000\n",
      "L = 19.178521, acc = 0.950000\n",
      "L = 19.161564, acc = 0.950000\n",
      "L = 19.144640, acc = 0.950000\n",
      "L = 19.127749, acc = 0.950000\n",
      "L = 19.110892, acc = 0.950000\n",
      "L = 19.094068, acc = 0.950000\n",
      "L = 19.077277, acc = 0.950000\n",
      "L = 19.060520, acc = 0.950000\n",
      "L = 19.043795, acc = 0.950000\n",
      "L = 19.027104, acc = 0.950000\n",
      "L = 19.010446, acc = 0.950000\n",
      "L = 18.993821, acc = 0.950000\n",
      "L = 18.977229, acc = 0.950000\n",
      "L = 18.960670, acc = 0.950000\n",
      "L = 18.944145, acc = 0.950000\n",
      "L = 18.927652, acc = 0.950000\n",
      "L = 18.911192, acc = 0.950000\n",
      "L = 18.894765, acc = 0.950000\n",
      "L = 18.878371, acc = 0.950000\n",
      "L = 18.862010, acc = 0.950000\n",
      "L = 18.845682, acc = 0.950000\n",
      "L = 18.829387, acc = 0.955000\n",
      "L = 18.813125, acc = 0.955000\n",
      "L = 18.796895, acc = 0.955000\n",
      "L = 18.780699, acc = 0.955000\n",
      "L = 18.764535, acc = 0.955000\n",
      "L = 18.748403, acc = 0.955000\n",
      "L = 18.732305, acc = 0.955000\n",
      "L = 18.716239, acc = 0.955000\n",
      "L = 18.700206, acc = 0.955000\n",
      "L = 18.684205, acc = 0.955000\n",
      "L = 18.668237, acc = 0.955000\n",
      "L = 18.652302, acc = 0.955000\n",
      "L = 18.636399, acc = 0.955000\n",
      "L = 18.620529, acc = 0.955000\n",
      "L = 18.604691, acc = 0.955000\n",
      "L = 18.588885, acc = 0.955000\n",
      "L = 18.573112, acc = 0.955000\n",
      "L = 18.557371, acc = 0.955000\n",
      "L = 18.541663, acc = 0.955000\n",
      "L = 18.525987, acc = 0.955000\n",
      "L = 18.510343, acc = 0.955000\n",
      "L = 18.494731, acc = 0.955000\n",
      "L = 18.479152, acc = 0.955000\n",
      "L = 18.463605, acc = 0.955000\n",
      "L = 18.448089, acc = 0.955000\n",
      "L = 18.432606, acc = 0.955000\n",
      "L = 18.417155, acc = 0.955000\n",
      "L = 18.401736, acc = 0.955000\n",
      "L = 18.386349, acc = 0.955000\n",
      "L = 18.370994, acc = 0.955000\n",
      "L = 18.355671, acc = 0.955000\n",
      "L = 18.340379, acc = 0.955000\n",
      "L = 18.325120, acc = 0.955000\n",
      "L = 18.309892, acc = 0.955000\n",
      "L = 18.294696, acc = 0.955000\n",
      "L = 18.279532, acc = 0.955000\n",
      "L = 18.264399, acc = 0.955000\n",
      "L = 18.249298, acc = 0.955000\n",
      "L = 18.234228, acc = 0.955000\n",
      "L = 18.219190, acc = 0.955000\n",
      "L = 18.204184, acc = 0.955000\n",
      "L = 18.189209, acc = 0.955000\n",
      "L = 18.174265, acc = 0.955000\n",
      "L = 18.159353, acc = 0.955000\n",
      "L = 18.144472, acc = 0.955000\n",
      "L = 18.129623, acc = 0.955000\n",
      "L = 18.114804, acc = 0.955000\n",
      "L = 18.100017, acc = 0.955000\n",
      "L = 18.085261, acc = 0.955000\n",
      "L = 18.070536, acc = 0.955000\n",
      "L = 18.055842, acc = 0.955000\n",
      "L = 18.041179, acc = 0.955000\n",
      "L = 18.026547, acc = 0.955000\n",
      "L = 18.011946, acc = 0.955000\n",
      "L = 17.997376, acc = 0.955000\n",
      "L = 17.982837, acc = 0.955000\n",
      "L = 17.968328, acc = 0.955000\n",
      "L = 17.953851, acc = 0.955000\n",
      "L = 17.939403, acc = 0.955000\n",
      "L = 17.924987, acc = 0.955000\n",
      "L = 17.910601, acc = 0.955000\n",
      "L = 17.896246, acc = 0.955000\n",
      "L = 17.881921, acc = 0.955000\n",
      "L = 17.867627, acc = 0.955000\n",
      "L = 17.853363, acc = 0.955000\n",
      "L = 17.839129, acc = 0.955000\n",
      "L = 17.824926, acc = 0.955000\n",
      "L = 17.810753, acc = 0.955000\n",
      "L = 17.796610, acc = 0.955000\n",
      "L = 17.782497, acc = 0.955000\n",
      "L = 17.768415, acc = 0.955000\n",
      "L = 17.754362, acc = 0.955000\n",
      "L = 17.740340, acc = 0.955000\n",
      "L = 17.726347, acc = 0.955000\n",
      "L = 17.712384, acc = 0.955000\n",
      "L = 17.698452, acc = 0.955000\n",
      "L = 17.684549, acc = 0.955000\n",
      "L = 17.670675, acc = 0.955000\n",
      "L = 17.656832, acc = 0.955000\n",
      "L = 17.643018, acc = 0.955000\n",
      "L = 17.629233, acc = 0.955000\n",
      "L = 17.615479, acc = 0.955000\n",
      "L = 17.601753, acc = 0.955000\n",
      "L = 17.588057, acc = 0.955000\n",
      "L = 17.574391, acc = 0.955000\n",
      "L = 17.560754, acc = 0.955000\n",
      "L = 17.547146, acc = 0.955000\n",
      "L = 17.533567, acc = 0.955000\n",
      "L = 17.520018, acc = 0.955000\n",
      "L = 17.506497, acc = 0.955000\n",
      "L = 17.493006, acc = 0.955000\n",
      "L = 17.479543, acc = 0.955000\n",
      "L = 17.466110, acc = 0.955000\n",
      "L = 17.452705, acc = 0.955000\n",
      "L = 17.439330, acc = 0.955000\n",
      "L = 17.425983, acc = 0.955000\n",
      "L = 17.412665, acc = 0.955000\n",
      "L = 17.399375, acc = 0.955000\n",
      "L = 17.386114, acc = 0.955000\n",
      "L = 17.372882, acc = 0.955000\n",
      "L = 17.359678, acc = 0.955000\n",
      "L = 17.346503, acc = 0.955000\n",
      "L = 17.333356, acc = 0.955000\n",
      "L = 17.320237, acc = 0.955000\n",
      "L = 17.307147, acc = 0.955000\n",
      "L = 17.294085, acc = 0.955000\n",
      "L = 17.281051, acc = 0.955000\n",
      "L = 17.268045, acc = 0.955000\n",
      "L = 17.255068, acc = 0.955000\n",
      "L = 17.242118, acc = 0.955000\n",
      "L = 17.229196, acc = 0.955000\n",
      "L = 17.216302, acc = 0.955000\n",
      "L = 17.203436, acc = 0.955000\n",
      "L = 17.190598, acc = 0.955000\n",
      "L = 17.177787, acc = 0.955000\n",
      "L = 17.165004, acc = 0.955000\n",
      "L = 17.152249, acc = 0.955000\n",
      "L = 17.139521, acc = 0.955000\n",
      "L = 17.126821, acc = 0.955000\n",
      "L = 17.114148, acc = 0.955000\n",
      "L = 17.101502, acc = 0.955000\n",
      "L = 17.088884, acc = 0.955000\n",
      "L = 17.076293, acc = 0.955000\n",
      "L = 17.063730, acc = 0.955000\n",
      "L = 17.051193, acc = 0.955000\n",
      "L = 17.038683, acc = 0.955000\n",
      "L = 17.026201, acc = 0.955000\n",
      "L = 17.013745, acc = 0.955000\n",
      "L = 17.001317, acc = 0.955000\n",
      "L = 16.988915, acc = 0.955000\n",
      "L = 16.976540, acc = 0.955000\n",
      "L = 16.964192, acc = 0.955000\n",
      "L = 16.951870, acc = 0.955000\n",
      "L = 16.939575, acc = 0.955000\n",
      "L = 16.927307, acc = 0.955000\n",
      "L = 16.915065, acc = 0.955000\n",
      "L = 16.902850, acc = 0.955000\n",
      "L = 16.890661, acc = 0.955000\n",
      "L = 16.878498, acc = 0.955000\n",
      "L = 16.866361, acc = 0.955000\n",
      "L = 16.854251, acc = 0.955000\n",
      "L = 16.842167, acc = 0.955000\n",
      "L = 16.830109, acc = 0.955000\n",
      "L = 16.818077, acc = 0.955000\n",
      "L = 16.806071, acc = 0.955000\n",
      "L = 16.794091, acc = 0.955000\n",
      "L = 16.782136, acc = 0.955000\n",
      "L = 16.770208, acc = 0.955000\n",
      "L = 16.758305, acc = 0.955000\n",
      "L = 16.746428, acc = 0.955000\n",
      "L = 16.734576, acc = 0.955000\n",
      "L = 16.722750, acc = 0.955000\n",
      "L = 16.710950, acc = 0.955000\n",
      "L = 16.699175, acc = 0.955000\n",
      "L = 16.687425, acc = 0.955000\n",
      "L = 16.675701, acc = 0.955000\n",
      "L = 16.664002, acc = 0.955000\n",
      "L = 16.652328, acc = 0.955000\n",
      "L = 16.640679, acc = 0.955000\n",
      "L = 16.629056, acc = 0.955000\n",
      "L = 16.617457, acc = 0.955000\n",
      "L = 16.605883, acc = 0.955000\n",
      "L = 16.594334, acc = 0.955000\n",
      "L = 16.582811, acc = 0.955000\n",
      "L = 16.571311, acc = 0.955000\n",
      "L = 16.559837, acc = 0.955000\n",
      "L = 16.548387, acc = 0.955000\n",
      "L = 16.536962, acc = 0.955000\n",
      "L = 16.525561, acc = 0.955000\n",
      "L = 16.514185, acc = 0.955000\n",
      "L = 16.502834, acc = 0.955000\n",
      "L = 16.491506, acc = 0.955000\n",
      "L = 16.480203, acc = 0.955000\n",
      "L = 16.468925, acc = 0.955000\n",
      "L = 16.457670, acc = 0.955000\n",
      "L = 16.446440, acc = 0.955000\n",
      "L = 16.435233, acc = 0.955000\n",
      "L = 16.424051, acc = 0.955000\n",
      "L = 16.412893, acc = 0.955000\n",
      "L = 16.401758, acc = 0.955000\n",
      "L = 16.390648, acc = 0.955000\n",
      "L = 16.379561, acc = 0.955000\n",
      "L = 16.368498, acc = 0.955000\n",
      "L = 16.357458, acc = 0.955000\n",
      "L = 16.346442, acc = 0.955000\n",
      "L = 16.335450, acc = 0.955000\n",
      "L = 16.324481, acc = 0.955000\n",
      "L = 16.313536, acc = 0.955000\n",
      "L = 16.302614, acc = 0.955000\n",
      "L = 16.291715, acc = 0.955000\n",
      "L = 16.280840, acc = 0.955000\n",
      "L = 16.269987, acc = 0.955000\n",
      "L = 16.259158, acc = 0.955000\n",
      "L = 16.248352, acc = 0.955000\n",
      "L = 16.237569, acc = 0.955000\n",
      "L = 16.226809, acc = 0.955000\n",
      "L = 16.216072, acc = 0.955000\n",
      "L = 16.205358, acc = 0.955000\n",
      "L = 16.194666, acc = 0.955000\n",
      "L = 16.183998, acc = 0.955000\n",
      "L = 16.173352, acc = 0.955000\n",
      "L = 16.162728, acc = 0.955000\n",
      "L = 16.152127, acc = 0.955000\n",
      "L = 16.141549, acc = 0.955000\n",
      "L = 16.130993, acc = 0.960000\n",
      "L = 16.120459, acc = 0.960000\n",
      "L = 16.109948, acc = 0.960000\n",
      "L = 16.099459, acc = 0.960000\n",
      "L = 16.088993, acc = 0.960000\n",
      "L = 16.078548, acc = 0.960000\n",
      "L = 16.068126, acc = 0.960000\n",
      "L = 16.057725, acc = 0.960000\n",
      "L = 16.047347, acc = 0.960000\n",
      "L = 16.036990, acc = 0.960000\n",
      "L = 16.026656, acc = 0.960000\n",
      "L = 16.016343, acc = 0.960000\n",
      "L = 16.006052, acc = 0.960000\n",
      "L = 15.995782, acc = 0.960000\n",
      "L = 15.985535, acc = 0.960000\n",
      "L = 15.975309, acc = 0.960000\n",
      "L = 15.965104, acc = 0.960000\n",
      "L = 15.954921, acc = 0.960000\n",
      "L = 15.944759, acc = 0.960000\n",
      "L = 15.934619, acc = 0.960000\n",
      "L = 15.924500, acc = 0.960000\n",
      "L = 15.914402, acc = 0.960000\n",
      "L = 15.904326, acc = 0.960000\n",
      "L = 15.894271, acc = 0.960000\n",
      "L = 15.884236, acc = 0.960000\n",
      "L = 15.874223, acc = 0.960000\n",
      "L = 15.864231, acc = 0.960000\n",
      "L = 15.854260, acc = 0.960000\n",
      "L = 15.844309, acc = 0.960000\n",
      "L = 15.834379, acc = 0.960000\n",
      "L = 15.824471, acc = 0.960000\n",
      "L = 15.814582, acc = 0.960000\n",
      "L = 15.804715, acc = 0.960000\n",
      "L = 15.794868, acc = 0.960000\n",
      "L = 15.785042, acc = 0.960000\n",
      "L = 15.775236, acc = 0.960000\n",
      "L = 15.765450, acc = 0.960000\n",
      "L = 15.755685, acc = 0.960000\n",
      "L = 15.745940, acc = 0.960000\n",
      "L = 15.736216, acc = 0.960000\n",
      "L = 15.726511, acc = 0.960000\n",
      "L = 15.716827, acc = 0.960000\n",
      "L = 15.707163, acc = 0.960000\n",
      "L = 15.697519, acc = 0.960000\n",
      "L = 15.687895, acc = 0.960000\n",
      "L = 15.678291, acc = 0.960000\n",
      "L = 15.668707, acc = 0.960000\n",
      "L = 15.659143, acc = 0.960000\n",
      "L = 15.649598, acc = 0.960000\n",
      "L = 15.640074, acc = 0.960000\n",
      "L = 15.630569, acc = 0.960000\n",
      "L = 15.621083, acc = 0.960000\n",
      "L = 15.611617, acc = 0.960000\n",
      "L = 15.602171, acc = 0.960000\n",
      "L = 15.592744, acc = 0.960000\n",
      "L = 15.583336, acc = 0.960000\n",
      "L = 15.573948, acc = 0.960000\n",
      "L = 15.564579, acc = 0.960000\n",
      "L = 15.555230, acc = 0.960000\n",
      "L = 15.545900, acc = 0.960000\n",
      "L = 15.536588, acc = 0.960000\n",
      "L = 15.527296, acc = 0.960000\n",
      "L = 15.518023, acc = 0.960000\n",
      "L = 15.508769, acc = 0.960000\n",
      "L = 15.499534, acc = 0.960000\n",
      "L = 15.490318, acc = 0.960000\n",
      "L = 15.481121, acc = 0.960000\n",
      "L = 15.471942, acc = 0.960000\n",
      "L = 15.462782, acc = 0.960000\n",
      "L = 15.453641, acc = 0.960000\n",
      "L = 15.444519, acc = 0.960000\n",
      "L = 15.435415, acc = 0.960000\n",
      "L = 15.426329, acc = 0.960000\n",
      "L = 15.417263, acc = 0.960000\n",
      "L = 15.408214, acc = 0.960000\n",
      "L = 15.399184, acc = 0.960000\n",
      "L = 15.390173, acc = 0.960000\n",
      "L = 15.381179, acc = 0.960000\n",
      "L = 15.372204, acc = 0.960000\n",
      "L = 15.363247, acc = 0.960000\n",
      "L = 15.354308, acc = 0.960000\n",
      "L = 15.345388, acc = 0.960000\n",
      "L = 15.336485, acc = 0.960000\n",
      "L = 15.327600, acc = 0.960000\n",
      "L = 15.318733, acc = 0.960000\n",
      "L = 15.309885, acc = 0.960000\n",
      "L = 15.301054, acc = 0.960000\n",
      "L = 15.292240, acc = 0.960000\n",
      "L = 15.283445, acc = 0.960000\n",
      "L = 15.274667, acc = 0.960000\n",
      "L = 15.265907, acc = 0.960000\n",
      "L = 15.257165, acc = 0.960000\n",
      "L = 15.248440, acc = 0.960000\n",
      "L = 15.239732, acc = 0.960000\n",
      "L = 15.231042, acc = 0.960000\n",
      "L = 15.222369, acc = 0.960000\n",
      "L = 15.213714, acc = 0.960000\n",
      "L = 15.205076, acc = 0.960000\n",
      "L = 15.196456, acc = 0.960000\n",
      "L = 15.187852, acc = 0.960000\n",
      "L = 15.179266, acc = 0.960000\n",
      "L = 15.170697, acc = 0.960000\n",
      "L = 15.162145, acc = 0.960000\n",
      "L = 15.153610, acc = 0.960000\n",
      "L = 15.145092, acc = 0.960000\n",
      "L = 15.136590, acc = 0.960000\n",
      "L = 15.128106, acc = 0.960000\n",
      "L = 15.119639, acc = 0.960000\n",
      "L = 15.111188, acc = 0.960000\n",
      "L = 15.102754, acc = 0.960000\n",
      "L = 15.094337, acc = 0.960000\n",
      "L = 15.085937, acc = 0.960000\n",
      "L = 15.077553, acc = 0.960000\n",
      "L = 15.069186, acc = 0.960000\n",
      "L = 15.060835, acc = 0.960000\n",
      "L = 15.052500, acc = 0.960000\n",
      "L = 15.044183, acc = 0.960000\n",
      "L = 15.035881, acc = 0.960000\n",
      "L = 15.027596, acc = 0.960000\n",
      "L = 15.019327, acc = 0.960000\n",
      "L = 15.011074, acc = 0.960000\n",
      "L = 15.002838, acc = 0.960000\n",
      "L = 14.994618, acc = 0.960000\n",
      "L = 14.986413, acc = 0.960000\n",
      "L = 14.978225, acc = 0.960000\n",
      "L = 14.970053, acc = 0.960000\n",
      "L = 14.961897, acc = 0.960000\n",
      "L = 14.953757, acc = 0.960000\n",
      "L = 14.945633, acc = 0.960000\n",
      "L = 14.937524, acc = 0.960000\n",
      "L = 14.929432, acc = 0.960000\n",
      "L = 14.921355, acc = 0.960000\n",
      "L = 14.913294, acc = 0.960000\n",
      "L = 14.905248, acc = 0.960000\n",
      "L = 14.897218, acc = 0.960000\n",
      "L = 14.889204, acc = 0.960000\n",
      "L = 14.881205, acc = 0.960000\n",
      "L = 14.873222, acc = 0.960000\n",
      "L = 14.865254, acc = 0.960000\n",
      "L = 14.857302, acc = 0.960000\n",
      "L = 14.849365, acc = 0.960000\n",
      "L = 14.841443, acc = 0.960000\n",
      "L = 14.833537, acc = 0.960000\n",
      "L = 14.825646, acc = 0.960000\n",
      "L = 14.817770, acc = 0.960000\n",
      "L = 14.809909, acc = 0.960000\n",
      "L = 14.802063, acc = 0.960000\n",
      "L = 14.794233, acc = 0.960000\n",
      "L = 14.786417, acc = 0.960000\n",
      "L = 14.778617, acc = 0.960000\n",
      "L = 14.770831, acc = 0.960000\n",
      "L = 14.763060, acc = 0.960000\n",
      "L = 14.755304, acc = 0.960000\n",
      "L = 14.747563, acc = 0.960000\n",
      "L = 14.739837, acc = 0.960000\n",
      "L = 14.732126, acc = 0.960000\n",
      "L = 14.724429, acc = 0.960000\n",
      "L = 14.716747, acc = 0.960000\n",
      "L = 14.709079, acc = 0.960000\n",
      "L = 14.701426, acc = 0.960000\n",
      "L = 14.693788, acc = 0.960000\n",
      "L = 14.686164, acc = 0.960000\n",
      "L = 14.678554, acc = 0.960000\n",
      "L = 14.670959, acc = 0.960000\n",
      "L = 14.663379, acc = 0.960000\n",
      "L = 14.655813, acc = 0.960000\n",
      "L = 14.648261, acc = 0.960000\n",
      "L = 14.640723, acc = 0.960000\n",
      "L = 14.633199, acc = 0.960000\n",
      "L = 14.625690, acc = 0.960000\n",
      "L = 14.618195, acc = 0.960000\n",
      "L = 14.610714, acc = 0.960000\n",
      "L = 14.603247, acc = 0.960000\n",
      "L = 14.595794, acc = 0.960000\n",
      "L = 14.588355, acc = 0.960000\n",
      "L = 14.580929, acc = 0.960000\n",
      "L = 14.573518, acc = 0.960000\n",
      "L = 14.566121, acc = 0.960000\n",
      "L = 14.558737, acc = 0.960000\n",
      "L = 14.551368, acc = 0.960000\n",
      "L = 14.544012, acc = 0.960000\n",
      "L = 14.536670, acc = 0.960000\n",
      "L = 14.529341, acc = 0.960000\n",
      "L = 14.522026, acc = 0.960000\n",
      "L = 14.514725, acc = 0.960000\n",
      "L = 14.507437, acc = 0.960000\n",
      "L = 14.500163, acc = 0.960000\n",
      "L = 14.492902, acc = 0.960000\n",
      "L = 14.485655, acc = 0.960000\n",
      "L = 14.478421, acc = 0.960000\n",
      "L = 14.471201, acc = 0.960000\n",
      "L = 14.463993, acc = 0.960000\n",
      "L = 14.456800, acc = 0.960000\n",
      "L = 14.449619, acc = 0.960000\n",
      "L = 14.442452, acc = 0.960000\n",
      "L = 14.435298, acc = 0.960000\n",
      "L = 14.428157, acc = 0.960000\n",
      "L = 14.421029, acc = 0.960000\n",
      "L = 14.413914, acc = 0.960000\n",
      "L = 14.406812, acc = 0.960000\n",
      "L = 14.399723, acc = 0.960000\n",
      "L = 14.392648, acc = 0.960000\n",
      "L = 14.385585, acc = 0.960000\n",
      "L = 14.378535, acc = 0.960000\n",
      "L = 14.371498, acc = 0.960000\n",
      "L = 14.364474, acc = 0.960000\n",
      "L = 14.357462, acc = 0.960000\n",
      "L = 14.350464, acc = 0.960000\n",
      "L = 14.343478, acc = 0.960000\n",
      "L = 14.336505, acc = 0.960000\n",
      "L = 14.329544, acc = 0.960000\n",
      "L = 14.322596, acc = 0.960000\n",
      "L = 14.315661, acc = 0.960000\n",
      "L = 14.308738, acc = 0.960000\n",
      "L = 14.301828, acc = 0.960000\n",
      "L = 14.294930, acc = 0.960000\n",
      "L = 14.288044, acc = 0.960000\n",
      "L = 14.281172, acc = 0.960000\n",
      "L = 14.274311, acc = 0.960000\n",
      "L = 14.267463, acc = 0.960000\n",
      "L = 14.260627, acc = 0.960000\n",
      "L = 14.253803, acc = 0.960000\n",
      "L = 14.246992, acc = 0.960000\n",
      "L = 14.240193, acc = 0.960000\n",
      "L = 14.233406, acc = 0.960000\n",
      "L = 14.226631, acc = 0.960000\n",
      "L = 14.219868, acc = 0.960000\n",
      "L = 14.213117, acc = 0.960000\n",
      "L = 14.206379, acc = 0.960000\n",
      "L = 14.199652, acc = 0.960000\n",
      "L = 14.192938, acc = 0.960000\n",
      "L = 14.186235, acc = 0.960000\n",
      "L = 14.179544, acc = 0.960000\n",
      "L = 14.172865, acc = 0.960000\n",
      "L = 14.166198, acc = 0.960000\n",
      "L = 14.159543, acc = 0.960000\n",
      "L = 14.152900, acc = 0.960000\n",
      "L = 14.146268, acc = 0.960000\n",
      "L = 14.139648, acc = 0.960000\n",
      "L = 14.133040, acc = 0.960000\n",
      "L = 14.126443, acc = 0.960000\n",
      "L = 14.119858, acc = 0.960000\n",
      "L = 14.113284, acc = 0.960000\n",
      "L = 14.106723, acc = 0.960000\n",
      "L = 14.100172, acc = 0.960000\n",
      "L = 14.093633, acc = 0.960000\n",
      "L = 14.087106, acc = 0.960000\n",
      "L = 14.080590, acc = 0.960000\n",
      "L = 14.074085, acc = 0.960000\n",
      "L = 14.067592, acc = 0.960000\n",
      "L = 14.061110, acc = 0.960000\n",
      "L = 14.054640, acc = 0.960000\n",
      "L = 14.048181, acc = 0.960000\n",
      "L = 14.041732, acc = 0.960000\n",
      "L = 14.035296, acc = 0.960000\n",
      "L = 14.028870, acc = 0.960000\n",
      "L = 14.022456, acc = 0.960000\n",
      "L = 14.016052, acc = 0.960000\n",
      "L = 14.009660, acc = 0.960000\n",
      "L = 14.003279, acc = 0.960000\n",
      "L = 13.996909, acc = 0.960000\n",
      "L = 13.990549, acc = 0.960000\n",
      "L = 13.984201, acc = 0.960000\n",
      "L = 13.977864, acc = 0.960000\n",
      "L = 13.971538, acc = 0.960000\n",
      "L = 13.965222, acc = 0.960000\n",
      "L = 13.958917, acc = 0.960000\n",
      "L = 13.952624, acc = 0.960000\n",
      "L = 13.946341, acc = 0.960000\n",
      "L = 13.940068, acc = 0.960000\n",
      "L = 13.933807, acc = 0.960000\n"
     ]
    }
   ],
   "source": [
    "# use the NN model and training\n",
    "nn = NN_Model([2, 6, 2])\n",
    "nn.init_weight()\n",
    "nn.backpropagation(X, t, 2000)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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": [
    "## 深入分析"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.02361974 0.97508818]\n",
      " [0.27772832 0.72007122]\n",
      " [0.83062658 0.17256861]\n",
      " [0.03436929 0.96345229]\n",
      " [0.02995351 0.96781939]\n",
      " [0.85808896 0.14486074]\n",
      " [0.05079962 0.94997004]\n",
      " [0.87805685 0.12525219]\n",
      " [0.03874122 0.96039457]]\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\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. 如何抽象，让神经网络的类支持更多的类型的层"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## References\n",
    "* 反向传播算法\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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}