autoencoder/CVAE_Experiment_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CVAE Experiment MNIST",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/manishiitg/ML_Experiments/blob/master/autoencoder/CVAE_Experiment_MNIST.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b657Qn4cRFnS",
        "colab_type": "text"
      },
      "source": [
        "# VAE Conditional Autoencoder"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KjsbslQAcSe5",
        "colab_type": "code",
        "outputId": "296f3964-4132-4aff-e247-263c78e8bc8b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torchvision  \n",
        "import torchvision.transforms as transforms\n",
        "from torchvision import datasets\n",
        "from torch.utils.data import DataLoader\n",
        "\n",
        "\n",
        "import numpy as np\n",
        "import torch.nn.functional as F\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline  \n",
        "\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "print(device)"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "cuda\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rl3pWfVwVSd-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "51a882ad-3b9b-4a39-f97f-e49e445376c7"
      },
      "source": [
        "import PIL.Image as Image\n",
        "\n",
        "\n",
        "transform = transforms.Compose(\n",
        "    [\n",
        "     transforms.ToTensor(),\n",
        "     transforms.Normalize((0.5,), (0.5,))\n",
        "    ])\n",
        "\n",
        "batch_size = 128\n",
        "mnist_data = torchvision.datasets.MNIST(\"data\", train=True,transform=transform, download=True)\n",
        "dataset_len = len(mnist_data)\n",
        "data_loader = DataLoader(mnist_data,batch_size=batch_size)\n"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\r0it [00:00, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to data/MNIST/raw/train-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "9920512it [00:01, 8605762.16it/s]                            \n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting data/MNIST/raw/train-images-idx3-ubyte.gz to data/MNIST/raw\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "  0%|          | 0/28881 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to data/MNIST/raw/train-labels-idx1-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "32768it [00:00, 141602.35it/s]           \n",
            "  0%|          | 0/1648877 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting data/MNIST/raw/train-labels-idx1-ubyte.gz to data/MNIST/raw\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to data/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "1654784it [00:00, 2311854.19it/s]                            \n",
            "  0%|          | 0/4542 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting data/MNIST/raw/t10k-images-idx3-ubyte.gz to data/MNIST/raw\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to data/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\r8192it [00:00, 53581.35it/s]            "
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting data/MNIST/raw/t10k-labels-idx1-ubyte.gz to data/MNIST/raw\n",
            "Processing...\n",
            "Done!\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FxN_CM5KcKdJ",
        "colab_type": "code",
        "outputId": "a3d947c8-070b-43a4-9f97-e09136c0952c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        }
      },
      "source": [
        "def show(img,renorm=False,nrow=8,interpolation='bicubic'):\n",
        "  if renorm:\n",
        "    img = img/2 + 0.5\n",
        "  img_grid = torchvision.utils.make_grid(img,nrow=nrow).numpy()\n",
        "  plt.figure()\n",
        "  plt.imshow(np.transpose(img_grid, (1,2,0)), interpolation=interpolation)\n",
        "  plt.axis('off')\n",
        "  plt.show()\n",
        "\n",
        "batch, _ = next(iter(data_loader))\n",
        "show(batch[0:64], renorm = True, nrow=8)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngwjiDUUwOIMI4g1FMDiDCOINxf8BupRdyFUMp1AAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mzQo1JcxYTIh",
        "colab_type": "text"
      },
      "source": [
        "Previewing the data "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q_cLElNcfS01",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class VAE(nn.Module):\n",
        "    def __init__(self, nc, ndf, ngf, z_dim=50, c_dims = 10):\n",
        "        super(VAE, self).__init__()\n",
        "        h_dim = nc * ndf * ngf\n",
        "        self.nc = nc\n",
        "        self.ndf = ndf\n",
        "        self.ngf = ngf\n",
        "        self.h_dim = h_dim\n",
        "        self.encoder = nn.Sequential(\n",
        "            nn.Linear(h_dim  + c_dims, 2000),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(2000, 500),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(500, 100),\n",
        "            nn.ReLU()\n",
        "        )   \n",
        "        \n",
        "        self.fc1 = nn.Linear(100, z_dim)\n",
        "        self.fc2 = nn.Linear(100, z_dim)\n",
        "        \n",
        "        self.decoder = nn.Sequential(\n",
        "            nn.Linear(z_dim  + c_dims, 100),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(100, 500),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(500, 2000),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(2000, h_dim),\n",
        "            nn.Sigmoid()\n",
        "        )\n",
        "        \n",
        "    def reparameterize(self, mu, logvar):\n",
        "        std = logvar.mul(0.5).exp_().to(device)\n",
        "        esp = torch.randn(*mu.size()).to(device)\n",
        "        z = mu + std * esp\n",
        "        return z\n",
        "    \n",
        "    def bottleneck(self, h):\n",
        "        mu, logvar = self.fc1(h), self.fc2(h)\n",
        "        z = self.reparameterize(mu, logvar)\n",
        "        return z, mu, logvar\n",
        "\n",
        "    def encode(self, x, c):\n",
        "        # print(\"encoder input shape\" , x.shape)\n",
        "        # print(\"conditional shape\", c.shape)\n",
        "        x = torch.cat([x, c ], 1)\n",
        "        h = self.encoder(x)\n",
        "        z, mu, logvar = self.bottleneck(h)\n",
        "        return z, mu, logvar\n",
        "\n",
        "    def decode(self, z,c ):\n",
        "        z = torch.cat([z, c ], 1)\n",
        "        z = self.decoder(z)\n",
        "        return z\n",
        "\n",
        "    def forward(self, x, c):\n",
        "        \n",
        "        # print(x.shape, \"input shape\")\n",
        "        x = x.view(-1, self.nc * self.ndf * self.ngf)\n",
        "        # print(\"reshape\", x.shape)\n",
        "        z, mu, logvar = self.encode(x,c)\n",
        "        # print(\"resampleing\" , z.shape)\n",
        "        z = self.decode(z, c)\n",
        "        # print(z.shape, \"output shape\")\n",
        "        z = z.view(-1, self.nc, self.ndf , self.ngf )\n",
        "        return z, mu, logvar\n",
        "\n",
        "\n",
        "conditional_dims = 10\n",
        "\n",
        "batch, _ = next(iter(data_loader))\n",
        "img = batch[1]\n",
        "\n",
        "channels = img.size(0)\n",
        "w, h = img.size(1) , img.size(2) \n",
        "\n",
        "nc = channels\n",
        "ndf = w\n",
        "ngf = h\n",
        "\n",
        "latent_dim = 20\n",
        "\n",
        "model = VAE(nc=nc, ngf=ngf, ndf=ndf,c_dims=conditional_dims, z_dim = latent_dim).to(device)\n",
        "optimizer = torch.optim.Adam(model.parameters(), lr=.0005)\n",
        "\n",
        "def loss_fn(recon_x, x, mu, logvar):\n",
        "    BCE = F.binary_cross_entropy(recon_x, x, reduction='sum')\n",
        "    # BCE = F.mse_loss(recon_x, x, size_average=False)\n",
        "\n",
        "    # see Appendix B from VAE paper:\n",
        "    # Kingma and Welling. Auto-Encoding Variational Bayes. ICLR, 2014\n",
        "    # 0.5 * sum(1 + log(sigma^2) - mu^2 - sigma^2)\n",
        "    KLD = -0.5 * torch.sum(1 + logvar - mu.pow(2) - logvar.exp())\n",
        "\n",
        "    return BCE + KLD\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "euiKRu-ZyhFv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_one_hot(batch_size, label):\n",
        "  label.unsqueeze_(-1)\n",
        "  one_hot = torch.FloatTensor(len(label), conditional_dims).zero_()\n",
        "  one_hot.scatter_(1, label,1)\n",
        "  one_hot = one_hot.to(device)\n",
        "  return one_hot"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1f6uvQjMDgtk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(epoch):\n",
        "    model.train()\n",
        "    train_loss = 0\n",
        "    count_data = 0\n",
        "    batch_count_data = 0\n",
        "    for batch_idx, (data,label) in enumerate(data_loader):\n",
        "               \n",
        "        one_hot = get_one_hot(len(label), label)\n",
        "        images = data.to(device)\n",
        "        optimizer.zero_grad()\n",
        "        \n",
        "        recon_batch, mu, log_var = model(images, one_hot)\n",
        "        loss = loss_fn(recon_batch, images, mu, log_var)\n",
        "        \n",
        "        loss.backward()\n",
        "        train_loss += loss.item()\n",
        "        optimizer.step()\n",
        "\n",
        "        count_data += len(data)\n",
        "        batch_count_data += len(data)\n",
        "        \n",
        "        if batch_idx % 10 == 0:\n",
        "            print('Train Epoch: {} [{}/{} ({:.0f}%)]\\tLoss: {:.6f}'.format(\n",
        "                epoch, batch_idx * len(data), dataset_len,\n",
        "                100. * batch_idx / len(data_loader), loss.item() / batch_count_data))\n",
        "            batch_count_data = 0\n",
        "    print('====> Epoch: {} Average loss: {:.4f}'.format(epoch, train_loss / count_data))\n",
        "\n",
        "    return train_loss / count_data\n",
        "\n",
        "\n",
        "# train(1)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QkYVgfoJYgBl",
        "colab_type": "text"
      },
      "source": [
        "Setting up the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RcVLmdsNzS2c",
        "colab_type": "code",
        "outputId": "00973814-083b-430d-c0a6-a65275d4d76e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "for i in range(50):\n",
        "  train(i)\n",
        "\n",
        "epoch = 50\n",
        "train_loss = 0\n",
        "test_loss = 0\n",
        "torch.save(model.state_dict(), './cvae/models/Epoch_{}_Train_loss_{:.4f}_Test_loss_{:.4f}.pth'.format(epoch, train_loss, test_loss))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train Epoch: 0 [0/60000 (0%)]\tLoss: 544.247803\n",
            "Train Epoch: 0 [1280/60000 (2%)]\tLoss: -875.883789\n",
            "Train Epoch: 0 [2560/60000 (4%)]\tLoss: -2809.138086\n",
            "Train Epoch: 0 [3840/60000 (6%)]\tLoss: -3866.390625\n",
            "Train Epoch: 0 [5120/60000 (9%)]\tLoss: -3609.001953\n",
            "Train Epoch: 0 [6400/60000 (11%)]\tLoss: -2916.343359\n",
            "Train Epoch: 0 [7680/60000 (13%)]\tLoss: -3554.987891\n",
            "Train Epoch: 0 [8960/60000 (15%)]\tLoss: -3438.623047\n",
            "Train Epoch: 0 [10240/60000 (17%)]\tLoss: -3011.305859\n",
            "Train Epoch: 0 [11520/60000 (19%)]\tLoss: -3026.544922\n",
            "Train Epoch: 0 [12800/60000 (21%)]\tLoss: -3340.040625\n",
            "Train Epoch: 0 [14080/60000 (23%)]\tLoss: -2908.716992\n",
            "Train Epoch: 0 [15360/60000 (26%)]\tLoss: -3680.950781\n",
            "Train Epoch: 0 [16640/60000 (28%)]\tLoss: -3437.018750\n",
            "Train Epoch: 0 [17920/60000 (30%)]\tLoss: -3473.913672\n",
            "Train Epoch: 0 [19200/60000 (32%)]\tLoss: -3108.372656\n",
            "Train Epoch: 0 [20480/60000 (34%)]\tLoss: -3182.807227\n",
            "Train Epoch: 0 [21760/60000 (36%)]\tLoss: -2797.205273\n",
            "Train Epoch: 0 [23040/60000 (38%)]\tLoss: -2545.803906\n",
            "Train Epoch: 0 [24320/60000 (41%)]\tLoss: -2489.741016\n",
            "Train Epoch: 0 [25600/60000 (43%)]\tLoss: -2275.349609\n",
            "Train Epoch: 0 [26880/60000 (45%)]\tLoss: -2314.164258\n",
            "Train Epoch: 0 [28160/60000 (47%)]\tLoss: -2182.883984\n",
            "Train Epoch: 0 [29440/60000 (49%)]\tLoss: -2024.028516\n",
            "Train Epoch: 0 [30720/60000 (51%)]\tLoss: -2069.444922\n",
            "Train Epoch: 0 [32000/60000 (53%)]\tLoss: -2215.859766\n",
            "Train Epoch: 0 [33280/60000 (55%)]\tLoss: -1961.729492\n",
            "Train Epoch: 0 [34560/60000 (58%)]\tLoss: -2126.676367\n",
            "Train Epoch: 0 [35840/60000 (60%)]\tLoss: -1947.315430\n",
            "Train Epoch: 0 [37120/60000 (62%)]\tLoss: -2049.341602\n",
            "Train Epoch: 0 [38400/60000 (64%)]\tLoss: -2101.685742\n",
            "Train Epoch: 0 [39680/60000 (66%)]\tLoss: -2065.071094\n",
            "Train Epoch: 0 [40960/60000 (68%)]\tLoss: -1972.650586\n",
            "Train Epoch: 0 [42240/60000 (70%)]\tLoss: -2086.634766\n",
            "Train Epoch: 0 [43520/60000 (72%)]\tLoss: -2151.912891\n",
            "Train Epoch: 0 [44800/60000 (75%)]\tLoss: -2091.670508\n",
            "Train Epoch: 0 [46080/60000 (77%)]\tLoss: -2197.598242\n",
            "Train Epoch: 0 [47360/60000 (79%)]\tLoss: -2184.805273\n",
            "Train Epoch: 0 [48640/60000 (81%)]\tLoss: -2134.563867\n",
            "Train Epoch: 0 [49920/60000 (83%)]\tLoss: -2235.877539\n",
            "Train Epoch: 0 [51200/60000 (85%)]\tLoss: -2111.095703\n",
            "Train Epoch: 0 [52480/60000 (87%)]\tLoss: -2302.757031\n",
            "Train Epoch: 0 [53760/60000 (90%)]\tLoss: -2244.872852\n",
            "Train Epoch: 0 [55040/60000 (92%)]\tLoss: -2215.418555\n",
            "Train Epoch: 0 [56320/60000 (94%)]\tLoss: -2350.481836\n",
            "Train Epoch: 0 [57600/60000 (96%)]\tLoss: -2180.467383\n",
            "Train Epoch: 0 [58880/60000 (98%)]\tLoss: -2332.886914\n",
            "====> Epoch: 0 Average loss: -25186.5038\n",
            "Train Epoch: 1 [0/60000 (0%)]\tLoss: -21098.419922\n",
            "Train Epoch: 1 [1280/60000 (2%)]\tLoss: -2368.781836\n",
            "Train Epoch: 1 [2560/60000 (4%)]\tLoss: -2233.804102\n",
            "Train Epoch: 1 [3840/60000 (6%)]\tLoss: -2259.917188\n",
            "Train Epoch: 1 [5120/60000 (9%)]\tLoss: -2260.673242\n",
            "Train Epoch: 1 [6400/60000 (11%)]\tLoss: -2014.729492\n",
            "Train Epoch: 1 [7680/60000 (13%)]\tLoss: -1854.517383\n",
            "Train Epoch: 1 [8960/60000 (15%)]\tLoss: -1826.063281\n",
            "Train Epoch: 1 [10240/60000 (17%)]\tLoss: -1935.075000\n",
            "Train Epoch: 1 [11520/60000 (19%)]\tLoss: -1833.414258\n",
            "Train Epoch: 1 [12800/60000 (21%)]\tLoss: -1811.572852\n",
            "Train Epoch: 1 [14080/60000 (23%)]\tLoss: -1852.193359\n",
            "Train Epoch: 1 [15360/60000 (26%)]\tLoss: -1899.609570\n",
            "Train Epoch: 1 [16640/60000 (28%)]\tLoss: -1847.541797\n",
            "Train Epoch: 1 [17920/60000 (30%)]\tLoss: -1973.256055\n",
            "Train Epoch: 1 [19200/60000 (32%)]\tLoss: -2059.383789\n",
            "Train Epoch: 1 [20480/60000 (34%)]\tLoss: -2086.474414\n",
            "Train Epoch: 1 [21760/60000 (36%)]\tLoss: -2115.480859\n",
            "Train Epoch: 1 [23040/60000 (38%)]\tLoss: -2066.590625\n",
            "Train Epoch: 1 [24320/60000 (41%)]\tLoss: -1881.654102\n",
            "Train Epoch: 1 [25600/60000 (43%)]\tLoss: -2052.418359\n",
            "Train Epoch: 1 [26880/60000 (45%)]\tLoss: -2051.984375\n",
            "Train Epoch: 1 [28160/60000 (47%)]\tLoss: -2065.235547\n",
            "Train Epoch: 1 [29440/60000 (49%)]\tLoss: -2027.016211\n",
            "Train Epoch: 1 [30720/60000 (51%)]\tLoss: -1983.526172\n",
            "Train Epoch: 1 [32000/60000 (53%)]\tLoss: -2043.738867\n",
            "Train Epoch: 1 [33280/60000 (55%)]\tLoss: -1964.553320\n",
            "Train Epoch: 1 [34560/60000 (58%)]\tLoss: -2081.624609\n",
            "Train Epoch: 1 [35840/60000 (60%)]\tLoss: -1967.281836\n",
            "Train Epoch: 1 [37120/60000 (62%)]\tLoss: -2147.562305\n",
            "Train Epoch: 1 [38400/60000 (64%)]\tLoss: -1990.441406\n",
            "Train Epoch: 1 [39680/60000 (66%)]\tLoss: -1989.903320\n",
            "Train Epoch: 1 [40960/60000 (68%)]\tLoss: -1938.786133\n",
            "Train Epoch: 1 [42240/60000 (70%)]\tLoss: -2024.670312\n",
            "Train Epoch: 1 [43520/60000 (72%)]\tLoss: -2056.510352\n",
            "Train Epoch: 1 [44800/60000 (75%)]\tLoss: -2024.755664\n",
            "Train Epoch: 1 [46080/60000 (77%)]\tLoss: -2034.744922\n",
            "Train Epoch: 1 [47360/60000 (79%)]\tLoss: -2084.189648\n",
            "Train Epoch: 1 [48640/60000 (81%)]\tLoss: -2002.920312\n",
            "Train Epoch: 1 [49920/60000 (83%)]\tLoss: -1995.967969\n",
            "Train Epoch: 1 [51200/60000 (85%)]\tLoss: -2134.997461\n",
            "Train Epoch: 1 [52480/60000 (87%)]\tLoss: -2012.674805\n",
            "Train Epoch: 1 [53760/60000 (90%)]\tLoss: -2023.567188\n",
            "Train Epoch: 1 [55040/60000 (92%)]\tLoss: -2170.151953\n",
            "Train Epoch: 1 [56320/60000 (94%)]\tLoss: -2148.273633\n",
            "Train Epoch: 1 [57600/60000 (96%)]\tLoss: -1959.405469\n",
            "Train Epoch: 1 [58880/60000 (98%)]\tLoss: -1970.717383\n",
            "====> Epoch: 1 Average loss: -20188.4433\n",
            "Train Epoch: 2 [0/60000 (0%)]\tLoss: -21894.025391\n",
            "Train Epoch: 2 [1280/60000 (2%)]\tLoss: -2229.253906\n",
            "Train Epoch: 2 [2560/60000 (4%)]\tLoss: -2083.109961\n",
            "Train Epoch: 2 [3840/60000 (6%)]\tLoss: -2181.143750\n",
            "Train Epoch: 2 [5120/60000 (9%)]\tLoss: -2170.930273\n",
            "Train Epoch: 2 [6400/60000 (11%)]\tLoss: -2111.482031\n",
            "Train Epoch: 2 [7680/60000 (13%)]\tLoss: -1894.475195\n",
            "Train Epoch: 2 [8960/60000 (15%)]\tLoss: -1930.870703\n",
            "Train Epoch: 2 [10240/60000 (17%)]\tLoss: -2106.217578\n",
            "Train Epoch: 2 [11520/60000 (19%)]\tLoss: -2012.312891\n",
            "Train Epoch: 2 [12800/60000 (21%)]\tLoss: -2020.197070\n",
            "Train Epoch: 2 [14080/60000 (23%)]\tLoss: -2072.058008\n",
            "Train Epoch: 2 [15360/60000 (26%)]\tLoss: -1892.147266\n",
            "Train Epoch: 2 [16640/60000 (28%)]\tLoss: -2122.583789\n",
            "Train Epoch: 2 [17920/60000 (30%)]\tLoss: -2008.355273\n",
            "Train Epoch: 2 [19200/60000 (32%)]\tLoss: -1998.390625\n",
            "Train Epoch: 2 [20480/60000 (34%)]\tLoss: -2062.911914\n",
            "Train Epoch: 2 [21760/60000 (36%)]\tLoss: -2136.646484\n",
            "Train Epoch: 2 [23040/60000 (38%)]\tLoss: -2185.756250\n",
            "Train Epoch: 2 [24320/60000 (41%)]\tLoss: -2064.517969\n",
            "Train Epoch: 2 [25600/60000 (43%)]\tLoss: -2197.803711\n",
            "Train Epoch: 2 [26880/60000 (45%)]\tLoss: -2099.062695\n",
            "Train Epoch: 2 [28160/60000 (47%)]\tLoss: -2141.618750\n",
            "Train Epoch: 2 [29440/60000 (49%)]\tLoss: -2086.983789\n",
            "Train Epoch: 2 [30720/60000 (51%)]\tLoss: -2107.752539\n",
            "Train Epoch: 2 [32000/60000 (53%)]\tLoss: -2077.625586\n",
            "Train Epoch: 2 [33280/60000 (55%)]\tLoss: -1948.452930\n",
            "Train Epoch: 2 [34560/60000 (58%)]\tLoss: -2147.118750\n",
            "Train Epoch: 2 [35840/60000 (60%)]\tLoss: -1977.393164\n",
            "Train Epoch: 2 [37120/60000 (62%)]\tLoss: -2264.787695\n",
            "Train Epoch: 2 [38400/60000 (64%)]\tLoss: -2058.056641\n",
            "Train Epoch: 2 [39680/60000 (66%)]\tLoss: -1954.932617\n",
            "Train Epoch: 2 [40960/60000 (68%)]\tLoss: -2035.269922\n",
            "Train Epoch: 2 [42240/60000 (70%)]\tLoss: -2012.059961\n",
            "Train Epoch: 2 [43520/60000 (72%)]\tLoss: -2179.529492\n",
            "Train Epoch: 2 [44800/60000 (75%)]\tLoss: -1922.537305\n",
            "Train Epoch: 2 [46080/60000 (77%)]\tLoss: -2071.775977\n",
            "Train Epoch: 2 [47360/60000 (79%)]\tLoss: -2028.557031\n",
            "Train Epoch: 2 [48640/60000 (81%)]\tLoss: -2073.666016\n",
            "Train Epoch: 2 [49920/60000 (83%)]\tLoss: -2010.631250\n",
            "Train Epoch: 2 [51200/60000 (85%)]\tLoss: -1989.047656\n",
            "Train Epoch: 2 [52480/60000 (87%)]\tLoss: -2199.149219\n",
            "Train Epoch: 2 [53760/60000 (90%)]\tLoss: -1936.716211\n",
            "Train Epoch: 2 [55040/60000 (92%)]\tLoss: -2166.331445\n",
            "Train Epoch: 2 [56320/60000 (94%)]\tLoss: -2054.295508\n",
            "Train Epoch: 2 [57600/60000 (96%)]\tLoss: -1980.991016\n",
            "Train Epoch: 2 [58880/60000 (98%)]\tLoss: -1936.625195\n",
            "====> Epoch: 2 Average loss: -20669.5749\n",
            "Train Epoch: 3 [0/60000 (0%)]\tLoss: -21128.177734\n",
            "Train Epoch: 3 [1280/60000 (2%)]\tLoss: -2164.962109\n",
            "Train Epoch: 3 [2560/60000 (4%)]\tLoss: -2247.082227\n",
            "Train Epoch: 3 [3840/60000 (6%)]\tLoss: -2204.592773\n",
            "Train Epoch: 3 [5120/60000 (9%)]\tLoss: -2137.395898\n",
            "Train Epoch: 3 [6400/60000 (11%)]\tLoss: -2293.176562\n",
            "Train Epoch: 3 [7680/60000 (13%)]\tLoss: -2077.931836\n",
            "Train Epoch: 3 [8960/60000 (15%)]\tLoss: -2016.928906\n",
            "Train Epoch: 3 [10240/60000 (17%)]\tLoss: -2150.760352\n",
            "Train Epoch: 3 [11520/60000 (19%)]\tLoss: -2031.914844\n",
            "Train Epoch: 3 [12800/60000 (21%)]\tLoss: -2016.777344\n",
            "Train Epoch: 3 [14080/60000 (23%)]\tLoss: -2001.156836\n",
            "Train Epoch: 3 [15360/60000 (26%)]\tLoss: -1959.902734\n",
            "Train Epoch: 3 [16640/60000 (28%)]\tLoss: -1990.943555\n",
            "Train Epoch: 3 [17920/60000 (30%)]\tLoss: -2097.777930\n",
            "Train Epoch: 3 [19200/60000 (32%)]\tLoss: -2108.128320\n",
            "Train Epoch: 3 [20480/60000 (34%)]\tLoss: -2056.767773\n",
            "Train Epoch: 3 [21760/60000 (36%)]\tLoss: -2063.913867\n",
            "Train Epoch: 3 [23040/60000 (38%)]\tLoss: -2044.961133\n",
            "Train Epoch: 3 [24320/60000 (41%)]\tLoss: -2036.498047\n",
            "Train Epoch: 3 [25600/60000 (43%)]\tLoss: -2116.761523\n",
            "Train Epoch: 3 [26880/60000 (45%)]\tLoss: -2084.914453\n",
            "Train Epoch: 3 [28160/60000 (47%)]\tLoss: -2095.359180\n",
            "Train Epoch: 3 [29440/60000 (49%)]\tLoss: -2216.741797\n",
            "Train Epoch: 3 [30720/60000 (51%)]\tLoss: -2043.941016\n",
            "Train Epoch: 3 [32000/60000 (53%)]\tLoss: -2069.318555\n",
            "Train Epoch: 3 [33280/60000 (55%)]\tLoss: -2079.095508\n",
            "Train Epoch: 3 [34560/60000 (58%)]\tLoss: -2087.390234\n",
            "Train Epoch: 3 [35840/60000 (60%)]\tLoss: -2131.694141\n",
            "Train Epoch: 3 [37120/60000 (62%)]\tLoss: -2382.475781\n",
            "Train Epoch: 3 [38400/60000 (64%)]\tLoss: -2024.423438\n",
            "Train Epoch: 3 [39680/60000 (66%)]\tLoss: -2095.900391\n",
            "Train Epoch: 3 [40960/60000 (68%)]\tLoss: -2034.214648\n",
            "Train Epoch: 3 [42240/60000 (70%)]\tLoss: -2071.925000\n",
            "Train Epoch: 3 [43520/60000 (72%)]\tLoss: -2086.831641\n",
            "Train Epoch: 3 [44800/60000 (75%)]\tLoss: -1990.118359\n",
            "Train Epoch: 3 [46080/60000 (77%)]\tLoss: -1971.611523\n",
            "Train Epoch: 3 [47360/60000 (79%)]\tLoss: -1975.813672\n",
            "Train Epoch: 3 [48640/60000 (81%)]\tLoss: -2116.305469\n",
            "Train Epoch: 3 [49920/60000 (83%)]\tLoss: -1996.817383\n",
            "Train Epoch: 3 [51200/60000 (85%)]\tLoss: -2056.414062\n",
            "Train Epoch: 3 [52480/60000 (87%)]\tLoss: -2145.407617\n",
            "Train Epoch: 3 [53760/60000 (90%)]\tLoss: -1963.461914\n",
            "Train Epoch: 3 [55040/60000 (92%)]\tLoss: -2051.337891\n",
            "Train Epoch: 3 [56320/60000 (94%)]\tLoss: -1990.978516\n",
            "Train Epoch: 3 [57600/60000 (96%)]\tLoss: -1981.485938\n",
            "Train Epoch: 3 [58880/60000 (98%)]\tLoss: -1919.848828\n",
            "====> Epoch: 3 Average loss: -20675.6634\n",
            "Train Epoch: 4 [0/60000 (0%)]\tLoss: -21482.585938\n",
            "Train Epoch: 4 [1280/60000 (2%)]\tLoss: -2266.546875\n",
            "Train Epoch: 4 [2560/60000 (4%)]\tLoss: -2179.568359\n",
            "Train Epoch: 4 [3840/60000 (6%)]\tLoss: -2056.721289\n",
            "Train Epoch: 4 [5120/60000 (9%)]\tLoss: -2114.875195\n",
            "Train Epoch: 4 [6400/60000 (11%)]\tLoss: -2054.142578\n",
            "Train Epoch: 4 [7680/60000 (13%)]\tLoss: -1984.795898\n",
            "Train Epoch: 4 [8960/60000 (15%)]\tLoss: -1992.910352\n",
            "Train Epoch: 4 [10240/60000 (17%)]\tLoss: -2044.080664\n",
            "Train Epoch: 4 [11520/60000 (19%)]\tLoss: -2024.383789\n",
            "Train Epoch: 4 [12800/60000 (21%)]\tLoss: -1951.407813\n",
            "Train Epoch: 4 [14080/60000 (23%)]\tLoss: -2173.644531\n",
            "Train Epoch: 4 [15360/60000 (26%)]\tLoss: -2003.282031\n",
            "Train Epoch: 4 [16640/60000 (28%)]\tLoss: -1999.455078\n",
            "Train Epoch: 4 [17920/60000 (30%)]\tLoss: -1985.379492\n",
            "Train Epoch: 4 [19200/60000 (32%)]\tLoss: -1970.292187\n",
            "Train Epoch: 4 [20480/60000 (34%)]\tLoss: -2065.288867\n",
            "Train Epoch: 4 [21760/60000 (36%)]\tLoss: -2082.311328\n",
            "Train Epoch: 4 [23040/60000 (38%)]\tLoss: -2063.572852\n",
            "Train Epoch: 4 [24320/60000 (41%)]\tLoss: -2000.800000\n",
            "Train Epoch: 4 [25600/60000 (43%)]\tLoss: -1968.251172\n",
            "Train Epoch: 4 [26880/60000 (45%)]\tLoss: -1993.044531\n",
            "Train Epoch: 4 [28160/60000 (47%)]\tLoss: -2093.256641\n",
            "Train Epoch: 4 [29440/60000 (49%)]\tLoss: -1999.972656\n",
            "Train Epoch: 4 [30720/60000 (51%)]\tLoss: -2005.656055\n",
            "Train Epoch: 4 [32000/60000 (53%)]\tLoss: -1982.630664\n",
            "Train Epoch: 4 [33280/60000 (55%)]\tLoss: -1936.140039\n",
            "Train Epoch: 4 [34560/60000 (58%)]\tLoss: -1966.029687\n",
            "Train Epoch: 4 [35840/60000 (60%)]\tLoss: -2036.001953\n",
            "Train Epoch: 4 [37120/60000 (62%)]\tLoss: -2248.260156\n",
            "Train Epoch: 4 [38400/60000 (64%)]\tLoss: -1946.291602\n",
            "Train Epoch: 4 [39680/60000 (66%)]\tLoss: -2033.114844\n",
            "Train Epoch: 4 [40960/60000 (68%)]\tLoss: -2024.277734\n",
            "Train Epoch: 4 [42240/60000 (70%)]\tLoss: -2073.887695\n",
            "Train Epoch: 4 [43520/60000 (72%)]\tLoss: -1981.513672\n",
            "Train Epoch: 4 [44800/60000 (75%)]\tLoss: -1936.252344\n",
            "Train Epoch: 4 [46080/60000 (77%)]\tLoss: -1977.712500\n",
            "Train Epoch: 4 [47360/60000 (79%)]\tLoss: -2020.864453\n",
            "Train Epoch: 4 [48640/60000 (81%)]\tLoss: -1974.442969\n",
            "Train Epoch: 4 [49920/60000 (83%)]\tLoss: -1995.123437\n",
            "Train Epoch: 4 [51200/60000 (85%)]\tLoss: -2047.208789\n",
            "Train Epoch: 4 [52480/60000 (87%)]\tLoss: -2101.730664\n",
            "Train Epoch: 4 [53760/60000 (90%)]\tLoss: -1950.947852\n",
            "Train Epoch: 4 [55040/60000 (92%)]\tLoss: -1992.666602\n",
            "Train Epoch: 4 [56320/60000 (94%)]\tLoss: -1992.716406\n",
            "Train Epoch: 4 [57600/60000 (96%)]\tLoss: -1935.474219\n",
            "Train Epoch: 4 [58880/60000 (98%)]\tLoss: -1964.483789\n",
            "====> Epoch: 4 Average loss: -20185.4666\n",
            "Train Epoch: 5 [0/60000 (0%)]\tLoss: -19468.509766\n",
            "Train Epoch: 5 [1280/60000 (2%)]\tLoss: -2175.959961\n",
            "Train Epoch: 5 [2560/60000 (4%)]\tLoss: -2017.231055\n",
            "Train Epoch: 5 [3840/60000 (6%)]\tLoss: -2030.510156\n",
            "Train Epoch: 5 [5120/60000 (9%)]\tLoss: -2023.362891\n",
            "Train Epoch: 5 [6400/60000 (11%)]\tLoss: -1981.178906\n",
            "Train Epoch: 5 [7680/60000 (13%)]\tLoss: -1962.483594\n",
            "Train Epoch: 5 [8960/60000 (15%)]\tLoss: -1985.386719\n",
            "Train Epoch: 5 [10240/60000 (17%)]\tLoss: -2101.605469\n",
            "Train Epoch: 5 [11520/60000 (19%)]\tLoss: -1946.575000\n",
            "Train Epoch: 5 [12800/60000 (21%)]\tLoss: -1951.076172\n",
            "Train Epoch: 5 [14080/60000 (23%)]\tLoss: -2138.135938\n",
            "Train Epoch: 5 [15360/60000 (26%)]\tLoss: -1980.266211\n",
            "Train Epoch: 5 [16640/60000 (28%)]\tLoss: -2044.473828\n",
            "Train Epoch: 5 [17920/60000 (30%)]\tLoss: -1929.246094\n",
            "Train Epoch: 5 [19200/60000 (32%)]\tLoss: -2098.147656\n",
            "Train Epoch: 5 [20480/60000 (34%)]\tLoss: -1977.203320\n",
            "Train Epoch: 5 [21760/60000 (36%)]\tLoss: -2031.681250\n",
            "Train Epoch: 5 [23040/60000 (38%)]\tLoss: -2005.204102\n",
            "Train Epoch: 5 [24320/60000 (41%)]\tLoss: -2051.591406\n",
            "Train Epoch: 5 [25600/60000 (43%)]\tLoss: -1946.242773\n",
            "Train Epoch: 5 [26880/60000 (45%)]\tLoss: -1945.046094\n",
            "Train Epoch: 5 [28160/60000 (47%)]\tLoss: -2114.450781\n",
            "Train Epoch: 5 [29440/60000 (49%)]\tLoss: -1999.134766\n",
            "Train Epoch: 5 [30720/60000 (51%)]\tLoss: -2020.923438\n",
            "Train Epoch: 5 [32000/60000 (53%)]\tLoss: -1910.621875\n",
            "Train Epoch: 5 [33280/60000 (55%)]\tLoss: -1926.071484\n",
            "Train Epoch: 5 [34560/60000 (58%)]\tLoss: -1941.067969\n",
            "Train Epoch: 5 [35840/60000 (60%)]\tLoss: -2048.529492\n",
            "Train Epoch: 5 [37120/60000 (62%)]\tLoss: -2162.010352\n",
            "Train Epoch: 5 [38400/60000 (64%)]\tLoss: -1937.491406\n",
            "Train Epoch: 5 [39680/60000 (66%)]\tLoss: -2003.472461\n",
            "Train Epoch: 5 [40960/60000 (68%)]\tLoss: -2038.576953\n",
            "Train Epoch: 5 [42240/60000 (70%)]\tLoss: -2040.864453\n",
            "Train Epoch: 5 [43520/60000 (72%)]\tLoss: -1954.669727\n",
            "Train Epoch: 5 [44800/60000 (75%)]\tLoss: -1946.348047\n",
            "Train Epoch: 5 [46080/60000 (77%)]\tLoss: -2011.774805\n",
            "Train Epoch: 5 [47360/60000 (79%)]\tLoss: -1975.217187\n",
            "Train Epoch: 5 [48640/60000 (81%)]\tLoss: -1993.572852\n",
            "Train Epoch: 5 [49920/60000 (83%)]\tLoss: -1959.112695\n",
            "Train Epoch: 5 [51200/60000 (85%)]\tLoss: -2021.993359\n",
            "Train Epoch: 5 [52480/60000 (87%)]\tLoss: -2081.356055\n",
            "Train Epoch: 5 [53760/60000 (90%)]\tLoss: -1977.585547\n",
            "Train Epoch: 5 [55040/60000 (92%)]\tLoss: -1997.437891\n",
            "Train Epoch: 5 [56320/60000 (94%)]\tLoss: -2008.827344\n",
            "Train Epoch: 5 [57600/60000 (96%)]\tLoss: -1943.475195\n",
            "Train Epoch: 5 [58880/60000 (98%)]\tLoss: -1906.105078\n",
            "====> Epoch: 5 Average loss: -19945.1537\n",
            "Train Epoch: 6 [0/60000 (0%)]\tLoss: -19914.828125\n",
            "Train Epoch: 6 [1280/60000 (2%)]\tLoss: -2209.577539\n",
            "Train Epoch: 6 [2560/60000 (4%)]\tLoss: -1954.771484\n",
            "Train Epoch: 6 [3840/60000 (6%)]\tLoss: -2002.598047\n",
            "Train Epoch: 6 [5120/60000 (9%)]\tLoss: -2040.839453\n",
            "Train Epoch: 6 [6400/60000 (11%)]\tLoss: -1950.550977\n",
            "Train Epoch: 6 [7680/60000 (13%)]\tLoss: -1979.857422\n",
            "Train Epoch: 6 [8960/60000 (15%)]\tLoss: -2023.302930\n",
            "Train Epoch: 6 [10240/60000 (17%)]\tLoss: -2103.297070\n",
            "Train Epoch: 6 [11520/60000 (19%)]\tLoss: -1970.880469\n",
            "Train Epoch: 6 [12800/60000 (21%)]\tLoss: -1950.102734\n",
            "Train Epoch: 6 [14080/60000 (23%)]\tLoss: -2234.721289\n",
            "Train Epoch: 6 [15360/60000 (26%)]\tLoss: -1969.202148\n",
            "Train Epoch: 6 [16640/60000 (28%)]\tLoss: -2016.502734\n",
            "Train Epoch: 6 [17920/60000 (30%)]\tLoss: -1943.241406\n",
            "Train Epoch: 6 [19200/60000 (32%)]\tLoss: -2135.690820\n",
            "Train Epoch: 6 [20480/60000 (34%)]\tLoss: -1966.687305\n",
            "Train Epoch: 6 [21760/60000 (36%)]\tLoss: -2071.983203\n",
            "Train Epoch: 6 [23040/60000 (38%)]\tLoss: -1983.699805\n",
            "Train Epoch: 6 [24320/60000 (41%)]\tLoss: -2027.255664\n",
            "Train Epoch: 6 [25600/60000 (43%)]\tLoss: -1987.824023\n",
            "Train Epoch: 6 [26880/60000 (45%)]\tLoss: -1976.623828\n",
            "Train Epoch: 6 [28160/60000 (47%)]\tLoss: -2103.829492\n",
            "Train Epoch: 6 [29440/60000 (49%)]\tLoss: -1953.107812\n",
            "Train Epoch: 6 [30720/60000 (51%)]\tLoss: -1935.510742\n",
            "Train Epoch: 6 [32000/60000 (53%)]\tLoss: -1928.831836\n",
            "Train Epoch: 6 [33280/60000 (55%)]\tLoss: -1916.837891\n",
            "Train Epoch: 6 [34560/60000 (58%)]\tLoss: -1927.520898\n",
            "Train Epoch: 6 [35840/60000 (60%)]\tLoss: -1989.501367\n",
            "Train Epoch: 6 [37120/60000 (62%)]\tLoss: -2115.278516\n",
            "Train Epoch: 6 [38400/60000 (64%)]\tLoss: -1953.514648\n",
            "Train Epoch: 6 [39680/60000 (66%)]\tLoss: -1963.070117\n",
            "Train Epoch: 6 [40960/60000 (68%)]\tLoss: -1941.978906\n",
            "Train Epoch: 6 [42240/60000 (70%)]\tLoss: -1970.448047\n",
            "Train Epoch: 6 [43520/60000 (72%)]\tLoss: -1941.456250\n",
            "Train Epoch: 6 [44800/60000 (75%)]\tLoss: -1969.330469\n",
            "Train Epoch: 6 [46080/60000 (77%)]\tLoss: -2003.152148\n",
            "Train Epoch: 6 [47360/60000 (79%)]\tLoss: -1984.823242\n",
            "Train Epoch: 6 [48640/60000 (81%)]\tLoss: -1907.383984\n",
            "Train Epoch: 6 [49920/60000 (83%)]\tLoss: -1961.769727\n",
            "Train Epoch: 6 [51200/60000 (85%)]\tLoss: -1958.219922\n",
            "Train Epoch: 6 [52480/60000 (87%)]\tLoss: -2035.358594\n",
            "Train Epoch: 6 [53760/60000 (90%)]\tLoss: -1984.889844\n",
            "Train Epoch: 6 [55040/60000 (92%)]\tLoss: -1940.711328\n",
            "Train Epoch: 6 [56320/60000 (94%)]\tLoss: -1990.954297\n",
            "Train Epoch: 6 [57600/60000 (96%)]\tLoss: -1931.316211\n",
            "Train Epoch: 6 [58880/60000 (98%)]\tLoss: -1913.616992\n",
            "====> Epoch: 6 Average loss: -19868.6260\n",
            "Train Epoch: 7 [0/60000 (0%)]\tLoss: -19490.292969\n",
            "Train Epoch: 7 [1280/60000 (2%)]\tLoss: -2146.627148\n",
            "Train Epoch: 7 [2560/60000 (4%)]\tLoss: -1938.744336\n",
            "Train Epoch: 7 [3840/60000 (6%)]\tLoss: -1990.894922\n",
            "Train Epoch: 7 [5120/60000 (9%)]\tLoss: -1990.659180\n",
            "Train Epoch: 7 [6400/60000 (11%)]\tLoss: -1944.448437\n",
            "Train Epoch: 7 [7680/60000 (13%)]\tLoss: -1958.376172\n",
            "Train Epoch: 7 [8960/60000 (15%)]\tLoss: -2000.036523\n",
            "Train Epoch: 7 [10240/60000 (17%)]\tLoss: -2109.914258\n",
            "Train Epoch: 7 [11520/60000 (19%)]\tLoss: -1958.632422\n",
            "Train Epoch: 7 [12800/60000 (21%)]\tLoss: -1947.502148\n",
            "Train Epoch: 7 [14080/60000 (23%)]\tLoss: -2134.345703\n",
            "Train Epoch: 7 [15360/60000 (26%)]\tLoss: -1951.000195\n",
            "Train Epoch: 7 [16640/60000 (28%)]\tLoss: -1961.305273\n",
            "Train Epoch: 7 [17920/60000 (30%)]\tLoss: -1957.655273\n",
            "Train Epoch: 7 [19200/60000 (32%)]\tLoss: -2080.166602\n",
            "Train Epoch: 7 [20480/60000 (34%)]\tLoss: -1969.685742\n",
            "Train Epoch: 7 [21760/60000 (36%)]\tLoss: -2006.977539\n",
            "Train Epoch: 7 [23040/60000 (38%)]\tLoss: -1975.934766\n",
            "Train Epoch: 7 [24320/60000 (41%)]\tLoss: -2019.415625\n",
            "Train Epoch: 7 [25600/60000 (43%)]\tLoss: -1956.698633\n",
            "Train Epoch: 7 [26880/60000 (45%)]\tLoss: -1948.833203\n",
            "Train Epoch: 7 [28160/60000 (47%)]\tLoss: -2010.691992\n",
            "Train Epoch: 7 [29440/60000 (49%)]\tLoss: -1954.517773\n",
            "Train Epoch: 7 [30720/60000 (51%)]\tLoss: -1970.550195\n",
            "Train Epoch: 7 [32000/60000 (53%)]\tLoss: -1923.856055\n",
            "Train Epoch: 7 [33280/60000 (55%)]\tLoss: -1931.949805\n",
            "Train Epoch: 7 [34560/60000 (58%)]\tLoss: -1938.038477\n",
            "Train Epoch: 7 [35840/60000 (60%)]\tLoss: -2000.940430\n",
            "Train Epoch: 7 [37120/60000 (62%)]\tLoss: -2068.551953\n",
            "Train Epoch: 7 [38400/60000 (64%)]\tLoss: -1958.182031\n",
            "Train Epoch: 7 [39680/60000 (66%)]\tLoss: -1978.781641\n",
            "Train Epoch: 7 [40960/60000 (68%)]\tLoss: -1983.950195\n",
            "Train Epoch: 7 [42240/60000 (70%)]\tLoss: -1995.607227\n",
            "Train Epoch: 7 [43520/60000 (72%)]\tLoss: -1945.805859\n",
            "Train Epoch: 7 [44800/60000 (75%)]\tLoss: -1971.181641\n",
            "Train Epoch: 7 [46080/60000 (77%)]\tLoss: -1984.672656\n",
            "Train Epoch: 7 [47360/60000 (79%)]\tLoss: -1979.578320\n",
            "Train Epoch: 7 [48640/60000 (81%)]\tLoss: -1894.412500\n",
            "Train Epoch: 7 [49920/60000 (83%)]\tLoss: -1981.979687\n",
            "Train Epoch: 7 [51200/60000 (85%)]\tLoss: -1957.245508\n",
            "Train Epoch: 7 [52480/60000 (87%)]\tLoss: -2026.048633\n",
            "Train Epoch: 7 [53760/60000 (90%)]\tLoss: -1964.015625\n",
            "Train Epoch: 7 [55040/60000 (92%)]\tLoss: -1944.688086\n",
            "Train Epoch: 7 [56320/60000 (94%)]\tLoss: -1967.947070\n",
            "Train Epoch: 7 [57600/60000 (96%)]\tLoss: -1938.541992\n",
            "Train Epoch: 7 [58880/60000 (98%)]\tLoss: -1912.716211\n",
            "====> Epoch: 7 Average loss: -19822.9860\n",
            "Train Epoch: 8 [0/60000 (0%)]\tLoss: -19686.236328\n",
            "Train Epoch: 8 [1280/60000 (2%)]\tLoss: -2081.908984\n",
            "Train Epoch: 8 [2560/60000 (4%)]\tLoss: -1959.465234\n",
            "Train Epoch: 8 [3840/60000 (6%)]\tLoss: -2002.221484\n",
            "Train Epoch: 8 [5120/60000 (9%)]\tLoss: -2007.152148\n",
            "Train Epoch: 8 [6400/60000 (11%)]\tLoss: -1952.549023\n",
            "Train Epoch: 8 [7680/60000 (13%)]\tLoss: -1960.817188\n",
            "Train Epoch: 8 [8960/60000 (15%)]\tLoss: -1983.225781\n",
            "Train Epoch: 8 [10240/60000 (17%)]\tLoss: -2145.461133\n",
            "Train Epoch: 8 [11520/60000 (19%)]\tLoss: -1940.282422\n",
            "Train Epoch: 8 [12800/60000 (21%)]\tLoss: -1955.590820\n",
            "Train Epoch: 8 [14080/60000 (23%)]\tLoss: -2127.726953\n",
            "Train Epoch: 8 [15360/60000 (26%)]\tLoss: -1958.531836\n",
            "Train Epoch: 8 [16640/60000 (28%)]\tLoss: -1964.050781\n",
            "Train Epoch: 8 [17920/60000 (30%)]\tLoss: -1963.077734\n",
            "Train Epoch: 8 [19200/60000 (32%)]\tLoss: -2039.068359\n",
            "Train Epoch: 8 [20480/60000 (34%)]\tLoss: -1956.649805\n",
            "Train Epoch: 8 [21760/60000 (36%)]\tLoss: -1986.917383\n",
            "Train Epoch: 8 [23040/60000 (38%)]\tLoss: -1968.595117\n",
            "Train Epoch: 8 [24320/60000 (41%)]\tLoss: -2016.834180\n",
            "Train Epoch: 8 [25600/60000 (43%)]\tLoss: -1962.879688\n",
            "Train Epoch: 8 [26880/60000 (45%)]\tLoss: -1963.497852\n",
            "Train Epoch: 8 [28160/60000 (47%)]\tLoss: -1987.444336\n",
            "Train Epoch: 8 [29440/60000 (49%)]\tLoss: -1962.837500\n",
            "Train Epoch: 8 [30720/60000 (51%)]\tLoss: -1961.876172\n",
            "Train Epoch: 8 [32000/60000 (53%)]\tLoss: -1936.542773\n",
            "Train Epoch: 8 [33280/60000 (55%)]\tLoss: -1937.327539\n",
            "Train Epoch: 8 [34560/60000 (58%)]\tLoss: -1939.633203\n",
            "Train Epoch: 8 [35840/60000 (60%)]\tLoss: -2006.457422\n",
            "Train Epoch: 8 [37120/60000 (62%)]\tLoss: -2057.271484\n",
            "Train Epoch: 8 [38400/60000 (64%)]\tLoss: -1966.019922\n",
            "Train Epoch: 8 [39680/60000 (66%)]\tLoss: -1973.222070\n",
            "Train Epoch: 8 [40960/60000 (68%)]\tLoss: -1984.141406\n",
            "Train Epoch: 8 [42240/60000 (70%)]\tLoss: -1986.711523\n",
            "Train Epoch: 8 [43520/60000 (72%)]\tLoss: -1954.401758\n",
            "Train Epoch: 8 [44800/60000 (75%)]\tLoss: -1976.390234\n",
            "Train Epoch: 8 [46080/60000 (77%)]\tLoss: -1989.461133\n",
            "Train Epoch: 8 [47360/60000 (79%)]\tLoss: -1992.135352\n",
            "Train Epoch: 8 [48640/60000 (81%)]\tLoss: -1912.782617\n",
            "Train Epoch: 8 [49920/60000 (83%)]\tLoss: -1975.372070\n",
            "Train Epoch: 8 [51200/60000 (85%)]\tLoss: -1989.336328\n",
            "Train Epoch: 8 [52480/60000 (87%)]\tLoss: -2033.026172\n",
            "Train Epoch: 8 [53760/60000 (90%)]\tLoss: -1946.962305\n",
            "Train Epoch: 8 [55040/60000 (92%)]\tLoss: -1963.321289\n",
            "Train Epoch: 8 [56320/60000 (94%)]\tLoss: -1969.944727\n",
            "Train Epoch: 8 [57600/60000 (96%)]\tLoss: -1940.954102\n",
            "Train Epoch: 8 [58880/60000 (98%)]\tLoss: -1920.922461\n",
            "====> Epoch: 8 Average loss: -19789.9925\n",
            "Train Epoch: 9 [0/60000 (0%)]\tLoss: -19555.019531\n",
            "Train Epoch: 9 [1280/60000 (2%)]\tLoss: -2025.910742\n",
            "Train Epoch: 9 [2560/60000 (4%)]\tLoss: -1964.202734\n",
            "Train Epoch: 9 [3840/60000 (6%)]\tLoss: -1981.232812\n",
            "Train Epoch: 9 [5120/60000 (9%)]\tLoss: -1962.315430\n",
            "Train Epoch: 9 [6400/60000 (11%)]\tLoss: -1947.405273\n",
            "Train Epoch: 9 [7680/60000 (13%)]\tLoss: -1972.027148\n",
            "Train Epoch: 9 [8960/60000 (15%)]\tLoss: -2005.300391\n",
            "Train Epoch: 9 [10240/60000 (17%)]\tLoss: -2146.231445\n",
            "Train Epoch: 9 [11520/60000 (19%)]\tLoss: -1936.803906\n",
            "Train Epoch: 9 [12800/60000 (21%)]\tLoss: -1952.432031\n",
            "Train Epoch: 9 [14080/60000 (23%)]\tLoss: -2066.446094\n",
            "Train Epoch: 9 [15360/60000 (26%)]\tLoss: -1959.095898\n",
            "Train Epoch: 9 [16640/60000 (28%)]\tLoss: -1965.474023\n",
            "Train Epoch: 9 [17920/60000 (30%)]\tLoss: -1946.262891\n",
            "Train Epoch: 9 [19200/60000 (32%)]\tLoss: -2017.856836\n",
            "Train Epoch: 9 [20480/60000 (34%)]\tLoss: -1940.588086\n",
            "Train Epoch: 9 [21760/60000 (36%)]\tLoss: -1967.974805\n",
            "Train Epoch: 9 [23040/60000 (38%)]\tLoss: -1977.028320\n",
            "Train Epoch: 9 [24320/60000 (41%)]\tLoss: -2034.087109\n",
            "Train Epoch: 9 [25600/60000 (43%)]\tLoss: -1924.162695\n",
            "Train Epoch: 9 [26880/60000 (45%)]\tLoss: -1953.834570\n",
            "Train Epoch: 9 [28160/60000 (47%)]\tLoss: -2020.530078\n",
            "Train Epoch: 9 [29440/60000 (49%)]\tLoss: -1950.614648\n",
            "Train Epoch: 9 [30720/60000 (51%)]\tLoss: -1956.832227\n",
            "Train Epoch: 9 [32000/60000 (53%)]\tLoss: -1937.833789\n",
            "Train Epoch: 9 [33280/60000 (55%)]\tLoss: -1935.237305\n",
            "Train Epoch: 9 [34560/60000 (58%)]\tLoss: -1939.011523\n",
            "Train Epoch: 9 [35840/60000 (60%)]\tLoss: -1990.620898\n",
            "Train Epoch: 9 [37120/60000 (62%)]\tLoss: -2048.708789\n",
            "Train Epoch: 9 [38400/60000 (64%)]\tLoss: -1975.557422\n",
            "Train Epoch: 9 [39680/60000 (66%)]\tLoss: -1967.772266\n",
            "Train Epoch: 9 [40960/60000 (68%)]\tLoss: -1985.258008\n",
            "Train Epoch: 9 [42240/60000 (70%)]\tLoss: -1989.835547\n",
            "Train Epoch: 9 [43520/60000 (72%)]\tLoss: -1953.613281\n",
            "Train Epoch: 9 [44800/60000 (75%)]\tLoss: -1959.430078\n",
            "Train Epoch: 9 [46080/60000 (77%)]\tLoss: -1986.797461\n",
            "Train Epoch: 9 [47360/60000 (79%)]\tLoss: -1955.259961\n",
            "Train Epoch: 9 [48640/60000 (81%)]\tLoss: -1914.529297\n",
            "Train Epoch: 9 [49920/60000 (83%)]\tLoss: -1974.772656\n",
            "Train Epoch: 9 [51200/60000 (85%)]\tLoss: -1976.551563\n",
            "Train Epoch: 9 [52480/60000 (87%)]\tLoss: -1996.225781\n",
            "Train Epoch: 9 [53760/60000 (90%)]\tLoss: -1968.327148\n",
            "Train Epoch: 9 [55040/60000 (92%)]\tLoss: -1937.938086\n",
            "Train Epoch: 9 [56320/60000 (94%)]\tLoss: -1964.672266\n",
            "Train Epoch: 9 [57600/60000 (96%)]\tLoss: -1937.560352\n",
            "Train Epoch: 9 [58880/60000 (98%)]\tLoss: -1918.775000\n",
            "====> Epoch: 9 Average loss: -19725.1648\n",
            "Train Epoch: 10 [0/60000 (0%)]\tLoss: -19824.490234\n",
            "Train Epoch: 10 [1280/60000 (2%)]\tLoss: -2133.915430\n",
            "Train Epoch: 10 [2560/60000 (4%)]\tLoss: -1926.991406\n",
            "Train Epoch: 10 [3840/60000 (6%)]\tLoss: -1979.789844\n",
            "Train Epoch: 10 [5120/60000 (9%)]\tLoss: -1950.213672\n",
            "Train Epoch: 10 [6400/60000 (11%)]\tLoss: -1957.757617\n",
            "Train Epoch: 10 [7680/60000 (13%)]\tLoss: -1982.273828\n",
            "Train Epoch: 10 [8960/60000 (15%)]\tLoss: -2001.548828\n",
            "Train Epoch: 10 [10240/60000 (17%)]\tLoss: -2149.199609\n",
            "Train Epoch: 10 [11520/60000 (19%)]\tLoss: -1935.204883\n",
            "Train Epoch: 10 [12800/60000 (21%)]\tLoss: -1953.814453\n",
            "Train Epoch: 10 [14080/60000 (23%)]\tLoss: -2059.950781\n",
            "Train Epoch: 10 [15360/60000 (26%)]\tLoss: -1963.485742\n",
            "Train Epoch: 10 [16640/60000 (28%)]\tLoss: -1947.428320\n",
            "Train Epoch: 10 [17920/60000 (30%)]\tLoss: -1996.299023\n",
            "Train Epoch: 10 [19200/60000 (32%)]\tLoss: -2008.135156\n",
            "Train Epoch: 10 [20480/60000 (34%)]\tLoss: -1950.858008\n",
            "Train Epoch: 10 [21760/60000 (36%)]\tLoss: -2004.602148\n",
            "Train Epoch: 10 [23040/60000 (38%)]\tLoss: -1964.979102\n",
            "Train Epoch: 10 [24320/60000 (41%)]\tLoss: -2029.787305\n",
            "Train Epoch: 10 [25600/60000 (43%)]\tLoss: -1960.151758\n",
            "Train Epoch: 10 [26880/60000 (45%)]\tLoss: -1955.189844\n",
            "Train Epoch: 10 [28160/60000 (47%)]\tLoss: -2064.230469\n",
            "Train Epoch: 10 [29440/60000 (49%)]\tLoss: -1944.205078\n",
            "Train Epoch: 10 [30720/60000 (51%)]\tLoss: -1960.329492\n",
            "Train Epoch: 10 [32000/60000 (53%)]\tLoss: -1913.708203\n",
            "Train Epoch: 10 [33280/60000 (55%)]\tLoss: -1956.313867\n",
            "Train Epoch: 10 [34560/60000 (58%)]\tLoss: -1936.780078\n",
            "Train Epoch: 10 [35840/60000 (60%)]\tLoss: -1999.389258\n",
            "Train Epoch: 10 [37120/60000 (62%)]\tLoss: -2046.830078\n",
            "Train Epoch: 10 [38400/60000 (64%)]\tLoss: -1981.725195\n",
            "Train Epoch: 10 [39680/60000 (66%)]\tLoss: -1983.907617\n",
            "Train Epoch: 10 [40960/60000 (68%)]\tLoss: -2004.258789\n",
            "Train Epoch: 10 [42240/60000 (70%)]\tLoss: -2011.806836\n",
            "Train Epoch: 10 [43520/60000 (72%)]\tLoss: -1946.136914\n",
            "Train Epoch: 10 [44800/60000 (75%)]\tLoss: -1974.413281\n",
            "Train Epoch: 10 [46080/60000 (77%)]\tLoss: -1966.410352\n",
            "Train Epoch: 10 [47360/60000 (79%)]\tLoss: -1951.403320\n",
            "Train Epoch: 10 [48640/60000 (81%)]\tLoss: -1939.028320\n",
            "Train Epoch: 10 [49920/60000 (83%)]\tLoss: -1985.716797\n",
            "Train Epoch: 10 [51200/60000 (85%)]\tLoss: -1986.153125\n",
            "Train Epoch: 10 [52480/60000 (87%)]\tLoss: -1975.980273\n",
            "Train Epoch: 10 [53760/60000 (90%)]\tLoss: -1970.391016\n",
            "Train Epoch: 10 [55040/60000 (92%)]\tLoss: -1948.287500\n",
            "Train Epoch: 10 [56320/60000 (94%)]\tLoss: -1967.335742\n",
            "Train Epoch: 10 [57600/60000 (96%)]\tLoss: -1945.053320\n",
            "Train Epoch: 10 [58880/60000 (98%)]\tLoss: -1922.739258\n",
            "====> Epoch: 10 Average loss: -19774.4638\n",
            "Train Epoch: 11 [0/60000 (0%)]\tLoss: -19621.009766\n",
            "Train Epoch: 11 [1280/60000 (2%)]\tLoss: -2114.725195\n",
            "Train Epoch: 11 [2560/60000 (4%)]\tLoss: -1951.466602\n",
            "Train Epoch: 11 [3840/60000 (6%)]\tLoss: -1998.380273\n",
            "Train Epoch: 11 [5120/60000 (9%)]\tLoss: -1925.284375\n",
            "Train Epoch: 11 [6400/60000 (11%)]\tLoss: -1962.659375\n",
            "Train Epoch: 11 [7680/60000 (13%)]\tLoss: -1947.879492\n",
            "Train Epoch: 11 [8960/60000 (15%)]\tLoss: -2014.585938\n",
            "Train Epoch: 11 [10240/60000 (17%)]\tLoss: -2154.150586\n",
            "Train Epoch: 11 [11520/60000 (19%)]\tLoss: -1936.534570\n",
            "Train Epoch: 11 [12800/60000 (21%)]\tLoss: -1957.584961\n",
            "Train Epoch: 11 [14080/60000 (23%)]\tLoss: -2111.470117\n",
            "Train Epoch: 11 [15360/60000 (26%)]\tLoss: -1973.163086\n",
            "Train Epoch: 11 [16640/60000 (28%)]\tLoss: -1952.662695\n",
            "Train Epoch: 11 [17920/60000 (30%)]\tLoss: -1958.252344\n",
            "Train Epoch: 11 [19200/60000 (32%)]\tLoss: -1976.790234\n",
            "Train Epoch: 11 [20480/60000 (34%)]\tLoss: -1949.942578\n",
            "Train Epoch: 11 [21760/60000 (36%)]\tLoss: -2049.282227\n",
            "Train Epoch: 11 [23040/60000 (38%)]\tLoss: -2009.258008\n",
            "Train Epoch: 11 [24320/60000 (41%)]\tLoss: -2021.148633\n",
            "Train Epoch: 11 [25600/60000 (43%)]\tLoss: -1962.112109\n",
            "Train Epoch: 11 [26880/60000 (45%)]\tLoss: -1979.532422\n",
            "Train Epoch: 11 [28160/60000 (47%)]\tLoss: -2003.060352\n",
            "Train Epoch: 11 [29440/60000 (49%)]\tLoss: -1951.819141\n",
            "Train Epoch: 11 [30720/60000 (51%)]\tLoss: -1971.577930\n",
            "Train Epoch: 11 [32000/60000 (53%)]\tLoss: -1925.903711\n",
            "Train Epoch: 11 [33280/60000 (55%)]\tLoss: -1960.315234\n",
            "Train Epoch: 11 [34560/60000 (58%)]\tLoss: -1937.726953\n",
            "Train Epoch: 11 [35840/60000 (60%)]\tLoss: -2001.833398\n",
            "Train Epoch: 11 [37120/60000 (62%)]\tLoss: -2048.210352\n",
            "Train Epoch: 11 [38400/60000 (64%)]\tLoss: -1985.798828\n",
            "Train Epoch: 11 [39680/60000 (66%)]\tLoss: -1987.502148\n",
            "Train Epoch: 11 [40960/60000 (68%)]\tLoss: -1996.734961\n",
            "Train Epoch: 11 [42240/60000 (70%)]\tLoss: -2014.971094\n",
            "Train Epoch: 11 [43520/60000 (72%)]\tLoss: -1949.638086\n",
            "Train Epoch: 11 [44800/60000 (75%)]\tLoss: -1975.660352\n",
            "Train Epoch: 11 [46080/60000 (77%)]\tLoss: -1973.657227\n",
            "Train Epoch: 11 [47360/60000 (79%)]\tLoss: -1969.258008\n",
            "Train Epoch: 11 [48640/60000 (81%)]\tLoss: -1978.711328\n",
            "Train Epoch: 11 [49920/60000 (83%)]\tLoss: -2019.863281\n",
            "Train Epoch: 11 [51200/60000 (85%)]\tLoss: -1979.044141\n",
            "Train Epoch: 11 [52480/60000 (87%)]\tLoss: -1978.175000\n",
            "Train Epoch: 11 [53760/60000 (90%)]\tLoss: -1967.957227\n",
            "Train Epoch: 11 [55040/60000 (92%)]\tLoss: -1950.036914\n",
            "Train Epoch: 11 [56320/60000 (94%)]\tLoss: -1980.920898\n",
            "Train Epoch: 11 [57600/60000 (96%)]\tLoss: -1950.010547\n",
            "Train Epoch: 11 [58880/60000 (98%)]\tLoss: -1919.280469\n",
            "====> Epoch: 11 Average loss: -19771.7038\n",
            "Train Epoch: 12 [0/60000 (0%)]\tLoss: -19545.378906\n",
            "Train Epoch: 12 [1280/60000 (2%)]\tLoss: -2083.675781\n",
            "Train Epoch: 12 [2560/60000 (4%)]\tLoss: -1973.000391\n",
            "Train Epoch: 12 [3840/60000 (6%)]\tLoss: -1999.550781\n",
            "Train Epoch: 12 [5120/60000 (9%)]\tLoss: -1907.886133\n",
            "Train Epoch: 12 [6400/60000 (11%)]\tLoss: -1968.886328\n",
            "Train Epoch: 12 [7680/60000 (13%)]\tLoss: -1964.722656\n",
            "Train Epoch: 12 [8960/60000 (15%)]\tLoss: -2003.379102\n",
            "Train Epoch: 12 [10240/60000 (17%)]\tLoss: -2143.312695\n",
            "Train Epoch: 12 [11520/60000 (19%)]\tLoss: -1934.255273\n",
            "Train Epoch: 12 [12800/60000 (21%)]\tLoss: -1961.786719\n",
            "Train Epoch: 12 [14080/60000 (23%)]\tLoss: -2130.649805\n",
            "Train Epoch: 12 [15360/60000 (26%)]\tLoss: -1970.591211\n",
            "Train Epoch: 12 [16640/60000 (28%)]\tLoss: -1952.067383\n",
            "Train Epoch: 12 [17920/60000 (30%)]\tLoss: -1948.497266\n",
            "Train Epoch: 12 [19200/60000 (32%)]\tLoss: -2013.159766\n",
            "Train Epoch: 12 [20480/60000 (34%)]\tLoss: -1939.954688\n",
            "Train Epoch: 12 [21760/60000 (36%)]\tLoss: -2021.011133\n",
            "Train Epoch: 12 [23040/60000 (38%)]\tLoss: -1997.648828\n",
            "Train Epoch: 12 [24320/60000 (41%)]\tLoss: -2024.914648\n",
            "Train Epoch: 12 [25600/60000 (43%)]\tLoss: -1936.856445\n",
            "Train Epoch: 12 [26880/60000 (45%)]\tLoss: -1984.312500\n",
            "Train Epoch: 12 [28160/60000 (47%)]\tLoss: -2031.229102\n",
            "Train Epoch: 12 [29440/60000 (49%)]\tLoss: -1956.297070\n",
            "Train Epoch: 12 [30720/60000 (51%)]\tLoss: -1963.159766\n",
            "Train Epoch: 12 [32000/60000 (53%)]\tLoss: -1929.481836\n",
            "Train Epoch: 12 [33280/60000 (55%)]\tLoss: -1957.760547\n",
            "Train Epoch: 12 [34560/60000 (58%)]\tLoss: -1928.083203\n",
            "Train Epoch: 12 [35840/60000 (60%)]\tLoss: -2010.052148\n",
            "Train Epoch: 12 [37120/60000 (62%)]\tLoss: -2053.409375\n",
            "Train Epoch: 12 [38400/60000 (64%)]\tLoss: -1990.806250\n",
            "Train Epoch: 12 [39680/60000 (66%)]\tLoss: -1995.315625\n",
            "Train Epoch: 12 [40960/60000 (68%)]\tLoss: -1995.313672\n",
            "Train Epoch: 12 [42240/60000 (70%)]\tLoss: -2005.049414\n",
            "Train Epoch: 12 [43520/60000 (72%)]\tLoss: -1951.809180\n",
            "Train Epoch: 12 [44800/60000 (75%)]\tLoss: -1977.198047\n",
            "Train Epoch: 12 [46080/60000 (77%)]\tLoss: -1994.112695\n",
            "Train Epoch: 12 [47360/60000 (79%)]\tLoss: -1964.942383\n",
            "Train Epoch: 12 [48640/60000 (81%)]\tLoss: -1932.419531\n",
            "Train Epoch: 12 [49920/60000 (83%)]\tLoss: -2015.223047\n",
            "Train Epoch: 12 [51200/60000 (85%)]\tLoss: -1984.416602\n",
            "Train Epoch: 12 [52480/60000 (87%)]\tLoss: -1980.862305\n",
            "Train Epoch: 12 [53760/60000 (90%)]\tLoss: -1968.525586\n",
            "Train Epoch: 12 [55040/60000 (92%)]\tLoss: -1954.887305\n",
            "Train Epoch: 12 [56320/60000 (94%)]\tLoss: -1971.972461\n",
            "Train Epoch: 12 [57600/60000 (96%)]\tLoss: -1949.892188\n",
            "Train Epoch: 12 [58880/60000 (98%)]\tLoss: -1924.068750\n",
            "====> Epoch: 12 Average loss: -19754.8631\n",
            "Train Epoch: 13 [0/60000 (0%)]\tLoss: -19554.669922\n",
            "Train Epoch: 13 [1280/60000 (2%)]\tLoss: -2101.539844\n",
            "Train Epoch: 13 [2560/60000 (4%)]\tLoss: -1968.387695\n",
            "Train Epoch: 13 [3840/60000 (6%)]\tLoss: -1985.440430\n",
            "Train Epoch: 13 [5120/60000 (9%)]\tLoss: -1914.148438\n",
            "Train Epoch: 13 [6400/60000 (11%)]\tLoss: -1951.634570\n",
            "Train Epoch: 13 [7680/60000 (13%)]\tLoss: -1964.501172\n",
            "Train Epoch: 13 [8960/60000 (15%)]\tLoss: -1996.361914\n",
            "Train Epoch: 13 [10240/60000 (17%)]\tLoss: -2156.065039\n",
            "Train Epoch: 13 [11520/60000 (19%)]\tLoss: -1940.483984\n",
            "Train Epoch: 13 [12800/60000 (21%)]\tLoss: -1963.062305\n",
            "Train Epoch: 13 [14080/60000 (23%)]\tLoss: -2088.174414\n",
            "Train Epoch: 13 [15360/60000 (26%)]\tLoss: -1976.578320\n",
            "Train Epoch: 13 [16640/60000 (28%)]\tLoss: -1954.293555\n",
            "Train Epoch: 13 [17920/60000 (30%)]\tLoss: -1945.507227\n",
            "Train Epoch: 13 [19200/60000 (32%)]\tLoss: -1995.187305\n",
            "Train Epoch: 13 [20480/60000 (34%)]\tLoss: -1947.766797\n",
            "Train Epoch: 13 [21760/60000 (36%)]\tLoss: -2021.284570\n",
            "Train Epoch: 13 [23040/60000 (38%)]\tLoss: -2016.320898\n",
            "Train Epoch: 13 [24320/60000 (41%)]\tLoss: -2024.027148\n",
            "Train Epoch: 13 [25600/60000 (43%)]\tLoss: -1959.657422\n",
            "Train Epoch: 13 [26880/60000 (45%)]\tLoss: -1985.427930\n",
            "Train Epoch: 13 [28160/60000 (47%)]\tLoss: -2011.813867\n",
            "Train Epoch: 13 [29440/60000 (49%)]\tLoss: -1958.359570\n",
            "Train Epoch: 13 [30720/60000 (51%)]\tLoss: -1976.196680\n",
            "Train Epoch: 13 [32000/60000 (53%)]\tLoss: -1916.926367\n",
            "Train Epoch: 13 [33280/60000 (55%)]\tLoss: -1939.724414\n",
            "Train Epoch: 13 [34560/60000 (58%)]\tLoss: -1940.090430\n",
            "Train Epoch: 13 [35840/60000 (60%)]\tLoss: -2007.244727\n",
            "Train Epoch: 13 [37120/60000 (62%)]\tLoss: -2081.513672\n",
            "Train Epoch: 13 [38400/60000 (64%)]\tLoss: -1989.782422\n",
            "Train Epoch: 13 [39680/60000 (66%)]\tLoss: -1988.036719\n",
            "Train Epoch: 13 [40960/60000 (68%)]\tLoss: -1995.552539\n",
            "Train Epoch: 13 [42240/60000 (70%)]\tLoss: -2002.985547\n",
            "Train Epoch: 13 [43520/60000 (72%)]\tLoss: -1953.429492\n",
            "Train Epoch: 13 [44800/60000 (75%)]\tLoss: -1972.611523\n",
            "Train Epoch: 13 [46080/60000 (77%)]\tLoss: -1987.545508\n",
            "Train Epoch: 13 [47360/60000 (79%)]\tLoss: -1948.539844\n",
            "Train Epoch: 13 [48640/60000 (81%)]\tLoss: -1928.389062\n",
            "Train Epoch: 13 [49920/60000 (83%)]\tLoss: -2002.379883\n",
            "Train Epoch: 13 [51200/60000 (85%)]\tLoss: -1996.974609\n",
            "Train Epoch: 13 [52480/60000 (87%)]\tLoss: -1982.738281\n",
            "Train Epoch: 13 [53760/60000 (90%)]\tLoss: -1966.525781\n",
            "Train Epoch: 13 [55040/60000 (92%)]\tLoss: -1963.196875\n",
            "Train Epoch: 13 [56320/60000 (94%)]\tLoss: -1964.599609\n",
            "Train Epoch: 13 [57600/60000 (96%)]\tLoss: -1952.077344\n",
            "Train Epoch: 13 [58880/60000 (98%)]\tLoss: -1921.858789\n",
            "====> Epoch: 13 Average loss: -19743.0136\n",
            "Train Epoch: 14 [0/60000 (0%)]\tLoss: -19570.847656\n",
            "Train Epoch: 14 [1280/60000 (2%)]\tLoss: -2138.163672\n",
            "Train Epoch: 14 [2560/60000 (4%)]\tLoss: -1960.586328\n",
            "Train Epoch: 14 [3840/60000 (6%)]\tLoss: -1964.523633\n",
            "Train Epoch: 14 [5120/60000 (9%)]\tLoss: -1917.086523\n",
            "Train Epoch: 14 [6400/60000 (11%)]\tLoss: -1952.333984\n",
            "Train Epoch: 14 [7680/60000 (13%)]\tLoss: -1968.330469\n",
            "Train Epoch: 14 [8960/60000 (15%)]\tLoss: -2011.784180\n",
            "Train Epoch: 14 [10240/60000 (17%)]\tLoss: -2098.408984\n",
            "Train Epoch: 14 [11520/60000 (19%)]\tLoss: -1931.804102\n",
            "Train Epoch: 14 [12800/60000 (21%)]\tLoss: -1989.414258\n",
            "Train Epoch: 14 [14080/60000 (23%)]\tLoss: -2101.643945\n",
            "Train Epoch: 14 [15360/60000 (26%)]\tLoss: -1972.979883\n",
            "Train Epoch: 14 [16640/60000 (28%)]\tLoss: -1957.203711\n",
            "Train Epoch: 14 [17920/60000 (30%)]\tLoss: -1982.518750\n",
            "Train Epoch: 14 [19200/60000 (32%)]\tLoss: -2010.047852\n",
            "Train Epoch: 14 [20480/60000 (34%)]\tLoss: -1942.766211\n",
            "Train Epoch: 14 [21760/60000 (36%)]\tLoss: -2035.315430\n",
            "Train Epoch: 14 [23040/60000 (38%)]\tLoss: -1964.907227\n",
            "Train Epoch: 14 [24320/60000 (41%)]\tLoss: -2027.541992\n",
            "Train Epoch: 14 [25600/60000 (43%)]\tLoss: -1952.895898\n",
            "Train Epoch: 14 [26880/60000 (45%)]\tLoss: -1983.464062\n",
            "Train Epoch: 14 [28160/60000 (47%)]\tLoss: -2022.414648\n",
            "Train Epoch: 14 [29440/60000 (49%)]\tLoss: -1961.027344\n",
            "Train Epoch: 14 [30720/60000 (51%)]\tLoss: -1969.208984\n",
            "Train Epoch: 14 [32000/60000 (53%)]\tLoss: -1931.599414\n",
            "Train Epoch: 14 [33280/60000 (55%)]\tLoss: -1939.193555\n",
            "Train Epoch: 14 [34560/60000 (58%)]\tLoss: -1922.522656\n",
            "Train Epoch: 14 [35840/60000 (60%)]\tLoss: -2008.243359\n",
            "Train Epoch: 14 [37120/60000 (62%)]\tLoss: -2043.532617\n",
            "Train Epoch: 14 [38400/60000 (64%)]\tLoss: -1991.187500\n",
            "Train Epoch: 14 [39680/60000 (66%)]\tLoss: -1990.210547\n",
            "Train Epoch: 14 [40960/60000 (68%)]\tLoss: -2008.587891\n",
            "Train Epoch: 14 [42240/60000 (70%)]\tLoss: -2000.163672\n",
            "Train Epoch: 14 [43520/60000 (72%)]\tLoss: -1952.607422\n",
            "Train Epoch: 14 [44800/60000 (75%)]\tLoss: -1985.216992\n",
            "Train Epoch: 14 [46080/60000 (77%)]\tLoss: -1990.969141\n",
            "Train Epoch: 14 [47360/60000 (79%)]\tLoss: -1972.135156\n",
            "Train Epoch: 14 [48640/60000 (81%)]\tLoss: -1934.403125\n",
            "Train Epoch: 14 [49920/60000 (83%)]\tLoss: -2008.271680\n",
            "Train Epoch: 14 [51200/60000 (85%)]\tLoss: -2001.427148\n",
            "Train Epoch: 14 [52480/60000 (87%)]\tLoss: -1948.127148\n",
            "Train Epoch: 14 [53760/60000 (90%)]\tLoss: -1969.553906\n",
            "Train Epoch: 14 [55040/60000 (92%)]\tLoss: -1943.136328\n",
            "Train Epoch: 14 [56320/60000 (94%)]\tLoss: -1966.717383\n",
            "Train Epoch: 14 [57600/60000 (96%)]\tLoss: -1952.197656\n",
            "Train Epoch: 14 [58880/60000 (98%)]\tLoss: -1921.660547\n",
            "====> Epoch: 14 Average loss: -19726.9058\n",
            "Train Epoch: 15 [0/60000 (0%)]\tLoss: -19379.884766\n",
            "Train Epoch: 15 [1280/60000 (2%)]\tLoss: -2086.852539\n",
            "Train Epoch: 15 [2560/60000 (4%)]\tLoss: -1957.832227\n",
            "Train Epoch: 15 [3840/60000 (6%)]\tLoss: -1952.599805\n",
            "Train Epoch: 15 [5120/60000 (9%)]\tLoss: -1918.512109\n",
            "Train Epoch: 15 [6400/60000 (11%)]\tLoss: -1946.244727\n",
            "Train Epoch: 15 [7680/60000 (13%)]\tLoss: -1972.146484\n",
            "Train Epoch: 15 [8960/60000 (15%)]\tLoss: -1989.558789\n",
            "Train Epoch: 15 [10240/60000 (17%)]\tLoss: -2081.512891\n",
            "Train Epoch: 15 [11520/60000 (19%)]\tLoss: -1930.482617\n",
            "Train Epoch: 15 [12800/60000 (21%)]\tLoss: -1967.272070\n",
            "Train Epoch: 15 [14080/60000 (23%)]\tLoss: -2114.193555\n",
            "Train Epoch: 15 [15360/60000 (26%)]\tLoss: -1972.878320\n",
            "Train Epoch: 15 [16640/60000 (28%)]\tLoss: -1954.064648\n",
            "Train Epoch: 15 [17920/60000 (30%)]\tLoss: -1951.637109\n",
            "Train Epoch: 15 [19200/60000 (32%)]\tLoss: -2030.820703\n",
            "Train Epoch: 15 [20480/60000 (34%)]\tLoss: -1942.928125\n",
            "Train Epoch: 15 [21760/60000 (36%)]\tLoss: -2026.775000\n",
            "Train Epoch: 15 [23040/60000 (38%)]\tLoss: -1985.248047\n",
            "Train Epoch: 15 [24320/60000 (41%)]\tLoss: -2028.907422\n",
            "Train Epoch: 15 [25600/60000 (43%)]\tLoss: -1923.175195\n",
            "Train Epoch: 15 [26880/60000 (45%)]\tLoss: -1967.093359\n",
            "Train Epoch: 15 [28160/60000 (47%)]\tLoss: -2020.504492\n",
            "Train Epoch: 15 [29440/60000 (49%)]\tLoss: -1960.282813\n",
            "Train Epoch: 15 [30720/60000 (51%)]\tLoss: -1970.307617\n",
            "Train Epoch: 15 [32000/60000 (53%)]\tLoss: -1936.507227\n",
            "Train Epoch: 15 [33280/60000 (55%)]\tLoss: -1937.407227\n",
            "Train Epoch: 15 [34560/60000 (58%)]\tLoss: -1921.004492\n",
            "Train Epoch: 15 [35840/60000 (60%)]\tLoss: -2011.411523\n",
            "Train Epoch: 15 [37120/60000 (62%)]\tLoss: -2090.379102\n",
            "Train Epoch: 15 [38400/60000 (64%)]\tLoss: -1991.155859\n",
            "Train Epoch: 15 [39680/60000 (66%)]\tLoss: -2006.966016\n",
            "Train Epoch: 15 [40960/60000 (68%)]\tLoss: -2004.792383\n",
            "Train Epoch: 15 [42240/60000 (70%)]\tLoss: -2010.104687\n",
            "Train Epoch: 15 [43520/60000 (72%)]\tLoss: -1957.071094\n",
            "Train Epoch: 15 [44800/60000 (75%)]\tLoss: -1976.277539\n",
            "Train Epoch: 15 [46080/60000 (77%)]\tLoss: -1989.374805\n",
            "Train Epoch: 15 [47360/60000 (79%)]\tLoss: -1964.082422\n",
            "Train Epoch: 15 [48640/60000 (81%)]\tLoss: -1934.186328\n",
            "Train Epoch: 15 [49920/60000 (83%)]\tLoss: -2007.164844\n",
            "Train Epoch: 15 [51200/60000 (85%)]\tLoss: -1960.325000\n",
            "Train Epoch: 15 [52480/60000 (87%)]\tLoss: -1961.841016\n",
            "Train Epoch: 15 [53760/60000 (90%)]\tLoss: -1968.444922\n",
            "Train Epoch: 15 [55040/60000 (92%)]\tLoss: -1943.097852\n",
            "Train Epoch: 15 [56320/60000 (94%)]\tLoss: -1966.204688\n",
            "Train Epoch: 15 [57600/60000 (96%)]\tLoss: -1954.356445\n",
            "Train Epoch: 15 [58880/60000 (98%)]\tLoss: -1918.953516\n",
            "====> Epoch: 15 Average loss: -19718.2222\n",
            "Train Epoch: 16 [0/60000 (0%)]\tLoss: -19366.238281\n",
            "Train Epoch: 16 [1280/60000 (2%)]\tLoss: -2107.650781\n",
            "Train Epoch: 16 [2560/60000 (4%)]\tLoss: -1936.945898\n",
            "Train Epoch: 16 [3840/60000 (6%)]\tLoss: -1972.564844\n",
            "Train Epoch: 16 [5120/60000 (9%)]\tLoss: -1937.219141\n",
            "Train Epoch: 16 [6400/60000 (11%)]\tLoss: -1944.154102\n",
            "Train Epoch: 16 [7680/60000 (13%)]\tLoss: -1971.633789\n",
            "Train Epoch: 16 [8960/60000 (15%)]\tLoss: -1980.591211\n",
            "Train Epoch: 16 [10240/60000 (17%)]\tLoss: -2063.220117\n",
            "Train Epoch: 16 [11520/60000 (19%)]\tLoss: -1946.166602\n",
            "Train Epoch: 16 [12800/60000 (21%)]\tLoss: -1968.882227\n",
            "Train Epoch: 16 [14080/60000 (23%)]\tLoss: -2099.528516\n",
            "Train Epoch: 16 [15360/60000 (26%)]\tLoss: -1966.364453\n",
            "Train Epoch: 16 [16640/60000 (28%)]\tLoss: -1954.928320\n",
            "Train Epoch: 16 [17920/60000 (30%)]\tLoss: -1949.946875\n",
            "Train Epoch: 16 [19200/60000 (32%)]\tLoss: -1965.453906\n",
            "Train Epoch: 16 [20480/60000 (34%)]\tLoss: -1928.983398\n",
            "Train Epoch: 16 [21760/60000 (36%)]\tLoss: -1984.216211\n",
            "Train Epoch: 16 [23040/60000 (38%)]\tLoss: -1963.579883\n",
            "Train Epoch: 16 [24320/60000 (41%)]\tLoss: -2014.899219\n",
            "Train Epoch: 16 [25600/60000 (43%)]\tLoss: -1938.487695\n",
            "Train Epoch: 16 [26880/60000 (45%)]\tLoss: -1967.276758\n",
            "Train Epoch: 16 [28160/60000 (47%)]\tLoss: -2000.405664\n",
            "Train Epoch: 16 [29440/60000 (49%)]\tLoss: -1939.642383\n",
            "Train Epoch: 16 [30720/60000 (51%)]\tLoss: -1971.165820\n",
            "Train Epoch: 16 [32000/60000 (53%)]\tLoss: -1908.761133\n",
            "Train Epoch: 16 [33280/60000 (55%)]\tLoss: -1939.189648\n",
            "Train Epoch: 16 [34560/60000 (58%)]\tLoss: -1922.736328\n",
            "Train Epoch: 16 [35840/60000 (60%)]\tLoss: -2007.786719\n",
            "Train Epoch: 16 [37120/60000 (62%)]\tLoss: -2033.338477\n",
            "Train Epoch: 16 [38400/60000 (64%)]\tLoss: -1967.956641\n",
            "Train Epoch: 16 [39680/60000 (66%)]\tLoss: -2008.460352\n",
            "Train Epoch: 16 [40960/60000 (68%)]\tLoss: -2000.141211\n",
            "Train Epoch: 16 [42240/60000 (70%)]\tLoss: -1970.566406\n",
            "Train Epoch: 16 [43520/60000 (72%)]\tLoss: -1955.106836\n",
            "Train Epoch: 16 [44800/60000 (75%)]\tLoss: -1970.813281\n",
            "Train Epoch: 16 [46080/60000 (77%)]\tLoss: -1968.965234\n",
            "Train Epoch: 16 [47360/60000 (79%)]\tLoss: -1961.461914\n",
            "Train Epoch: 16 [48640/60000 (81%)]\tLoss: -1952.358984\n",
            "Train Epoch: 16 [49920/60000 (83%)]\tLoss: -2005.946484\n",
            "Train Epoch: 16 [51200/60000 (85%)]\tLoss: -1982.372070\n",
            "Train Epoch: 16 [52480/60000 (87%)]\tLoss: -1957.127344\n",
            "Train Epoch: 16 [53760/60000 (90%)]\tLoss: -1968.555469\n",
            "Train Epoch: 16 [55040/60000 (92%)]\tLoss: -1943.544727\n",
            "Train Epoch: 16 [56320/60000 (94%)]\tLoss: -1965.956445\n",
            "Train Epoch: 16 [57600/60000 (96%)]\tLoss: -1951.375586\n",
            "Train Epoch: 16 [58880/60000 (98%)]\tLoss: -1921.673828\n",
            "====> Epoch: 16 Average loss: -19711.4418\n",
            "Train Epoch: 17 [0/60000 (0%)]\tLoss: -19304.855469\n",
            "Train Epoch: 17 [1280/60000 (2%)]\tLoss: -2115.795312\n",
            "Train Epoch: 17 [2560/60000 (4%)]\tLoss: -1952.259961\n",
            "Train Epoch: 17 [3840/60000 (6%)]\tLoss: -1966.235352\n",
            "Train Epoch: 17 [5120/60000 (9%)]\tLoss: -1920.550391\n",
            "Train Epoch: 17 [6400/60000 (11%)]\tLoss: -1946.964062\n",
            "Train Epoch: 17 [7680/60000 (13%)]\tLoss: -1971.521484\n",
            "Train Epoch: 17 [8960/60000 (15%)]\tLoss: -2022.439453\n",
            "Train Epoch: 17 [10240/60000 (17%)]\tLoss: -2106.335742\n",
            "Train Epoch: 17 [11520/60000 (19%)]\tLoss: -1936.634766\n",
            "Train Epoch: 17 [12800/60000 (21%)]\tLoss: -1972.404297\n",
            "Train Epoch: 17 [14080/60000 (23%)]\tLoss: -2096.025000\n",
            "Train Epoch: 17 [15360/60000 (26%)]\tLoss: -1969.126758\n",
            "Train Epoch: 17 [16640/60000 (28%)]\tLoss: -1955.209570\n",
            "Train Epoch: 17 [17920/60000 (30%)]\tLoss: -1949.737109\n",
            "Train Epoch: 17 [19200/60000 (32%)]\tLoss: -1987.873828\n",
            "Train Epoch: 17 [20480/60000 (34%)]\tLoss: -1928.459961\n",
            "Train Epoch: 17 [21760/60000 (36%)]\tLoss: -1990.148828\n",
            "Train Epoch: 17 [23040/60000 (38%)]\tLoss: -1971.977539\n",
            "Train Epoch: 17 [24320/60000 (41%)]\tLoss: -1997.207422\n",
            "Train Epoch: 17 [25600/60000 (43%)]\tLoss: -1933.923828\n",
            "Train Epoch: 17 [26880/60000 (45%)]\tLoss: -1970.985547\n",
            "Train Epoch: 17 [28160/60000 (47%)]\tLoss: -2000.375391\n",
            "Train Epoch: 17 [29440/60000 (49%)]\tLoss: -1945.430469\n",
            "Train Epoch: 17 [30720/60000 (51%)]\tLoss: -1959.218359\n",
            "Train Epoch: 17 [32000/60000 (53%)]\tLoss: -1940.222461\n",
            "Train Epoch: 17 [33280/60000 (55%)]\tLoss: -1943.094141\n",
            "Train Epoch: 17 [34560/60000 (58%)]\tLoss: -1931.006836\n",
            "Train Epoch: 17 [35840/60000 (60%)]\tLoss: -2008.034180\n",
            "Train Epoch: 17 [37120/60000 (62%)]\tLoss: -2045.376172\n",
            "Train Epoch: 17 [38400/60000 (64%)]\tLoss: -1987.395703\n",
            "Train Epoch: 17 [39680/60000 (66%)]\tLoss: -2007.254492\n",
            "Train Epoch: 17 [40960/60000 (68%)]\tLoss: -1991.348047\n",
            "Train Epoch: 17 [42240/60000 (70%)]\tLoss: -1997.080664\n",
            "Train Epoch: 17 [43520/60000 (72%)]\tLoss: -1946.713086\n",
            "Train Epoch: 17 [44800/60000 (75%)]\tLoss: -1960.664844\n",
            "Train Epoch: 17 [46080/60000 (77%)]\tLoss: -1968.103711\n",
            "Train Epoch: 17 [47360/60000 (79%)]\tLoss: -1955.200000\n",
            "Train Epoch: 17 [48640/60000 (81%)]\tLoss: -1948.508203\n",
            "Train Epoch: 17 [49920/60000 (83%)]\tLoss: -1984.555273\n",
            "Train Epoch: 17 [51200/60000 (85%)]\tLoss: -1992.582617\n",
            "Train Epoch: 17 [52480/60000 (87%)]\tLoss: -1949.404102\n",
            "Train Epoch: 17 [53760/60000 (90%)]\tLoss: -1967.444141\n",
            "Train Epoch: 17 [55040/60000 (92%)]\tLoss: -1946.622656\n",
            "Train Epoch: 17 [56320/60000 (94%)]\tLoss: -1967.958789\n",
            "Train Epoch: 17 [57600/60000 (96%)]\tLoss: -1951.280273\n",
            "Train Epoch: 17 [58880/60000 (98%)]\tLoss: -1925.901172\n",
            "====> Epoch: 17 Average loss: -19717.8441\n",
            "Train Epoch: 18 [0/60000 (0%)]\tLoss: -19344.876953\n",
            "Train Epoch: 18 [1280/60000 (2%)]\tLoss: -2074.729297\n",
            "Train Epoch: 18 [2560/60000 (4%)]\tLoss: -1951.353711\n",
            "Train Epoch: 18 [3840/60000 (6%)]\tLoss: -1966.900586\n",
            "Train Epoch: 18 [5120/60000 (9%)]\tLoss: -1940.786719\n",
            "Train Epoch: 18 [6400/60000 (11%)]\tLoss: -1948.715234\n",
            "Train Epoch: 18 [7680/60000 (13%)]\tLoss: -1951.348438\n",
            "Train Epoch: 18 [8960/60000 (15%)]\tLoss: -1986.385937\n",
            "Train Epoch: 18 [10240/60000 (17%)]\tLoss: -2094.278125\n",
            "Train Epoch: 18 [11520/60000 (19%)]\tLoss: -1946.274609\n",
            "Train Epoch: 18 [12800/60000 (21%)]\tLoss: -1949.225586\n",
            "Train Epoch: 18 [14080/60000 (23%)]\tLoss: -2090.169531\n",
            "Train Epoch: 18 [15360/60000 (26%)]\tLoss: -1968.447656\n",
            "Train Epoch: 18 [16640/60000 (28%)]\tLoss: -1954.672070\n",
            "Train Epoch: 18 [17920/60000 (30%)]\tLoss: -1973.299414\n",
            "Train Epoch: 18 [19200/60000 (32%)]\tLoss: -2016.327930\n",
            "Train Epoch: 18 [20480/60000 (34%)]\tLoss: -1922.628320\n",
            "Train Epoch: 18 [21760/60000 (36%)]\tLoss: -1983.394727\n",
            "Train Epoch: 18 [23040/60000 (38%)]\tLoss: -1974.695508\n",
            "Train Epoch: 18 [24320/60000 (41%)]\tLoss: -1992.604687\n",
            "Train Epoch: 18 [25600/60000 (43%)]\tLoss: -1937.037305\n",
            "Train Epoch: 18 [26880/60000 (45%)]\tLoss: -1971.573242\n",
            "Train Epoch: 18 [28160/60000 (47%)]\tLoss: -2006.856445\n",
            "Train Epoch: 18 [29440/60000 (49%)]\tLoss: -1950.394727\n",
            "Train Epoch: 18 [30720/60000 (51%)]\tLoss: -1970.570508\n",
            "Train Epoch: 18 [32000/60000 (53%)]\tLoss: -1926.273047\n",
            "Train Epoch: 18 [33280/60000 (55%)]\tLoss: -1941.337891\n",
            "Train Epoch: 18 [34560/60000 (58%)]\tLoss: -1922.762305\n",
            "Train Epoch: 18 [35840/60000 (60%)]\tLoss: -2000.290820\n",
            "Train Epoch: 18 [37120/60000 (62%)]\tLoss: -2037.032813\n",
            "Train Epoch: 18 [38400/60000 (64%)]\tLoss: -1972.234180\n",
            "Train Epoch: 18 [39680/60000 (66%)]\tLoss: -1989.661914\n",
            "Train Epoch: 18 [40960/60000 (68%)]\tLoss: -1984.003906\n",
            "Train Epoch: 18 [42240/60000 (70%)]\tLoss: -1975.878320\n",
            "Train Epoch: 18 [43520/60000 (72%)]\tLoss: -1937.796289\n",
            "Train Epoch: 18 [44800/60000 (75%)]\tLoss: -1972.856836\n",
            "Train Epoch: 18 [46080/60000 (77%)]\tLoss: -1972.675586\n",
            "Train Epoch: 18 [47360/60000 (79%)]\tLoss: -1952.715820\n",
            "Train Epoch: 18 [48640/60000 (81%)]\tLoss: -1968.176367\n",
            "Train Epoch: 18 [49920/60000 (83%)]\tLoss: -1985.271289\n",
            "Train Epoch: 18 [51200/60000 (85%)]\tLoss: -2007.177344\n",
            "Train Epoch: 18 [52480/60000 (87%)]\tLoss: -1944.841016\n",
            "Train Epoch: 18 [53760/60000 (90%)]\tLoss: -1969.836523\n",
            "Train Epoch: 18 [55040/60000 (92%)]\tLoss: -1939.539648\n",
            "Train Epoch: 18 [56320/60000 (94%)]\tLoss: -1964.119922\n",
            "Train Epoch: 18 [57600/60000 (96%)]\tLoss: -1953.203320\n",
            "Train Epoch: 18 [58880/60000 (98%)]\tLoss: -1925.526367\n",
            "====> Epoch: 18 Average loss: -19685.4689\n",
            "Train Epoch: 19 [0/60000 (0%)]\tLoss: -19270.546875\n",
            "Train Epoch: 19 [1280/60000 (2%)]\tLoss: -2055.898828\n",
            "Train Epoch: 19 [2560/60000 (4%)]\tLoss: -1950.667773\n",
            "Train Epoch: 19 [3840/60000 (6%)]\tLoss: -1944.044922\n",
            "Train Epoch: 19 [5120/60000 (9%)]\tLoss: -1939.366016\n",
            "Train Epoch: 19 [6400/60000 (11%)]\tLoss: -1949.600391\n",
            "Train Epoch: 19 [7680/60000 (13%)]\tLoss: -1969.190039\n",
            "Train Epoch: 19 [8960/60000 (15%)]\tLoss: -2014.926367\n",
            "Train Epoch: 19 [10240/60000 (17%)]\tLoss: -2101.387695\n",
            "Train Epoch: 19 [11520/60000 (19%)]\tLoss: -1965.964062\n",
            "Train Epoch: 19 [12800/60000 (21%)]\tLoss: -1950.431836\n",
            "Train Epoch: 19 [14080/60000 (23%)]\tLoss: -2094.220313\n",
            "Train Epoch: 19 [15360/60000 (26%)]\tLoss: -1968.353125\n",
            "Train Epoch: 19 [16640/60000 (28%)]\tLoss: -1951.120312\n",
            "Train Epoch: 19 [17920/60000 (30%)]\tLoss: -1969.412305\n",
            "Train Epoch: 19 [19200/60000 (32%)]\tLoss: -1988.522656\n",
            "Train Epoch: 19 [20480/60000 (34%)]\tLoss: -1922.109180\n",
            "Train Epoch: 19 [21760/60000 (36%)]\tLoss: -2011.688867\n",
            "Train Epoch: 19 [23040/60000 (38%)]\tLoss: -1978.191211\n",
            "Train Epoch: 19 [24320/60000 (41%)]\tLoss: -2016.610938\n",
            "Train Epoch: 19 [25600/60000 (43%)]\tLoss: -1919.481250\n",
            "Train Epoch: 19 [26880/60000 (45%)]\tLoss: -1972.802344\n",
            "Train Epoch: 19 [28160/60000 (47%)]\tLoss: -2011.926172\n",
            "Train Epoch: 19 [29440/60000 (49%)]\tLoss: -1943.938086\n",
            "Train Epoch: 19 [30720/60000 (51%)]\tLoss: -1970.353125\n",
            "Train Epoch: 19 [32000/60000 (53%)]\tLoss: -1907.723633\n",
            "Train Epoch: 19 [33280/60000 (55%)]\tLoss: -1932.558789\n",
            "Train Epoch: 19 [34560/60000 (58%)]\tLoss: -1923.212891\n",
            "Train Epoch: 19 [35840/60000 (60%)]\tLoss: -1991.703906\n",
            "Train Epoch: 19 [37120/60000 (62%)]\tLoss: -2050.243945\n",
            "Train Epoch: 19 [38400/60000 (64%)]\tLoss: -1977.782227\n",
            "Train Epoch: 19 [39680/60000 (66%)]\tLoss: -1993.563281\n",
            "Train Epoch: 19 [40960/60000 (68%)]\tLoss: -1993.642578\n",
            "Train Epoch: 19 [42240/60000 (70%)]\tLoss: -1984.585547\n",
            "Train Epoch: 19 [43520/60000 (72%)]\tLoss: -1935.572852\n",
            "Train Epoch: 19 [44800/60000 (75%)]\tLoss: -1981.561328\n",
            "Train Epoch: 19 [46080/60000 (77%)]\tLoss: -1968.046289\n",
            "Train Epoch: 19 [47360/60000 (79%)]\tLoss: -1952.988281\n",
            "Train Epoch: 19 [48640/60000 (81%)]\tLoss: -1948.274609\n",
            "Train Epoch: 19 [49920/60000 (83%)]\tLoss: -1976.208594\n",
            "Train Epoch: 19 [51200/60000 (85%)]\tLoss: -1972.998242\n",
            "Train Epoch: 19 [52480/60000 (87%)]\tLoss: -1957.855273\n",
            "Train Epoch: 19 [53760/60000 (90%)]\tLoss: -1969.167383\n",
            "Train Epoch: 19 [55040/60000 (92%)]\tLoss: -1941.436523\n",
            "Train Epoch: 19 [56320/60000 (94%)]\tLoss: -1964.439453\n",
            "Train Epoch: 19 [57600/60000 (96%)]\tLoss: -1949.237109\n",
            "Train Epoch: 19 [58880/60000 (98%)]\tLoss: -1930.196094\n",
            "====> Epoch: 19 Average loss: -19695.7441\n",
            "Train Epoch: 20 [0/60000 (0%)]\tLoss: -19307.492188\n",
            "Train Epoch: 20 [1280/60000 (2%)]\tLoss: -2086.527930\n",
            "Train Epoch: 20 [2560/60000 (4%)]\tLoss: -1951.227148\n",
            "Train Epoch: 20 [3840/60000 (6%)]\tLoss: -1938.319531\n",
            "Train Epoch: 20 [5120/60000 (9%)]\tLoss: -1918.656641\n",
            "Train Epoch: 20 [6400/60000 (11%)]\tLoss: -1945.885352\n",
            "Train Epoch: 20 [7680/60000 (13%)]\tLoss: -1970.763477\n",
            "Train Epoch: 20 [8960/60000 (15%)]\tLoss: -1990.945117\n",
            "Train Epoch: 20 [10240/60000 (17%)]\tLoss: -2068.758008\n",
            "Train Epoch: 20 [11520/60000 (19%)]\tLoss: -1955.244922\n",
            "Train Epoch: 20 [12800/60000 (21%)]\tLoss: -1948.523242\n",
            "Train Epoch: 20 [14080/60000 (23%)]\tLoss: -2104.286914\n",
            "Train Epoch: 20 [15360/60000 (26%)]\tLoss: -1963.827539\n",
            "Train Epoch: 20 [16640/60000 (28%)]\tLoss: -1953.754883\n",
            "Train Epoch: 20 [17920/60000 (30%)]\tLoss: -1973.662500\n",
            "Train Epoch: 20 [19200/60000 (32%)]\tLoss: -2017.347266\n",
            "Train Epoch: 20 [20480/60000 (34%)]\tLoss: -1925.046875\n",
            "Train Epoch: 20 [21760/60000 (36%)]\tLoss: -2027.263086\n",
            "Train Epoch: 20 [23040/60000 (38%)]\tLoss: -1972.859570\n",
            "Train Epoch: 20 [24320/60000 (41%)]\tLoss: -1957.130273\n",
            "Train Epoch: 20 [25600/60000 (43%)]\tLoss: -1917.892969\n",
            "Train Epoch: 20 [26880/60000 (45%)]\tLoss: -1970.304883\n",
            "Train Epoch: 20 [28160/60000 (47%)]\tLoss: -2014.603906\n",
            "Train Epoch: 20 [29440/60000 (49%)]\tLoss: -1939.263867\n",
            "Train Epoch: 20 [30720/60000 (51%)]\tLoss: -1972.985352\n",
            "Train Epoch: 20 [32000/60000 (53%)]\tLoss: -1930.415234\n",
            "Train Epoch: 20 [33280/60000 (55%)]\tLoss: -1935.761719\n",
            "Train Epoch: 20 [34560/60000 (58%)]\tLoss: -1926.960742\n",
            "Train Epoch: 20 [35840/60000 (60%)]\tLoss: -2005.334570\n",
            "Train Epoch: 20 [37120/60000 (62%)]\tLoss: -2043.042578\n",
            "Train Epoch: 20 [38400/60000 (64%)]\tLoss: -1967.025781\n",
            "Train Epoch: 20 [39680/60000 (66%)]\tLoss: -1984.422461\n",
            "Train Epoch: 20 [40960/60000 (68%)]\tLoss: -1995.699805\n",
            "Train Epoch: 20 [42240/60000 (70%)]\tLoss: -1973.890820\n",
            "Train Epoch: 20 [43520/60000 (72%)]\tLoss: -1931.213672\n",
            "Train Epoch: 20 [44800/60000 (75%)]\tLoss: -1976.098047\n",
            "Train Epoch: 20 [46080/60000 (77%)]\tLoss: -1980.893945\n",
            "Train Epoch: 20 [47360/60000 (79%)]\tLoss: -1949.909961\n",
            "Train Epoch: 20 [48640/60000 (81%)]\tLoss: -1942.331836\n",
            "Train Epoch: 20 [49920/60000 (83%)]\tLoss: -2004.747461\n",
            "Train Epoch: 20 [51200/60000 (85%)]\tLoss: -1996.858008\n",
            "Train Epoch: 20 [52480/60000 (87%)]\tLoss: -1970.604687\n",
            "Train Epoch: 20 [53760/60000 (90%)]\tLoss: -1969.223633\n",
            "Train Epoch: 20 [55040/60000 (92%)]\tLoss: -1941.061523\n",
            "Train Epoch: 20 [56320/60000 (94%)]\tLoss: -1967.682422\n",
            "Train Epoch: 20 [57600/60000 (96%)]\tLoss: -1945.083008\n",
            "Train Epoch: 20 [58880/60000 (98%)]\tLoss: -1929.932031\n",
            "====> Epoch: 20 Average loss: -19693.8685\n",
            "Train Epoch: 21 [0/60000 (0%)]\tLoss: -19511.261719\n",
            "Train Epoch: 21 [1280/60000 (2%)]\tLoss: -2070.198242\n",
            "Train Epoch: 21 [2560/60000 (4%)]\tLoss: -1947.439648\n",
            "Train Epoch: 21 [3840/60000 (6%)]\tLoss: -1939.314063\n",
            "Train Epoch: 21 [5120/60000 (9%)]\tLoss: -1928.779687\n",
            "Train Epoch: 21 [6400/60000 (11%)]\tLoss: -1942.175195\n",
            "Train Epoch: 21 [7680/60000 (13%)]\tLoss: -1955.026953\n",
            "Train Epoch: 21 [8960/60000 (15%)]\tLoss: -1980.927148\n",
            "Train Epoch: 21 [10240/60000 (17%)]\tLoss: -2104.497266\n",
            "Train Epoch: 21 [11520/60000 (19%)]\tLoss: -1934.517383\n",
            "Train Epoch: 21 [12800/60000 (21%)]\tLoss: -1946.646484\n",
            "Train Epoch: 21 [14080/60000 (23%)]\tLoss: -2113.191992\n",
            "Train Epoch: 21 [15360/60000 (26%)]\tLoss: -1963.586523\n",
            "Train Epoch: 21 [16640/60000 (28%)]\tLoss: -1954.353516\n",
            "Train Epoch: 21 [17920/60000 (30%)]\tLoss: -1969.713477\n",
            "Train Epoch: 21 [19200/60000 (32%)]\tLoss: -1998.988477\n",
            "Train Epoch: 21 [20480/60000 (34%)]\tLoss: -1921.978906\n",
            "Train Epoch: 21 [21760/60000 (36%)]\tLoss: -2000.058008\n",
            "Train Epoch: 21 [23040/60000 (38%)]\tLoss: -2002.967578\n",
            "Train Epoch: 21 [24320/60000 (41%)]\tLoss: -1986.243555\n",
            "Train Epoch: 21 [25600/60000 (43%)]\tLoss: -1936.103516\n",
            "Train Epoch: 21 [26880/60000 (45%)]\tLoss: -1961.446875\n",
            "Train Epoch: 21 [28160/60000 (47%)]\tLoss: -1999.540820\n",
            "Train Epoch: 21 [29440/60000 (49%)]\tLoss: -1949.326953\n",
            "Train Epoch: 21 [30720/60000 (51%)]\tLoss: -1973.751367\n",
            "Train Epoch: 21 [32000/60000 (53%)]\tLoss: -1925.237500\n",
            "Train Epoch: 21 [33280/60000 (55%)]\tLoss: -1934.535352\n",
            "Train Epoch: 21 [34560/60000 (58%)]\tLoss: -1927.103125\n",
            "Train Epoch: 21 [35840/60000 (60%)]\tLoss: -1981.501758\n",
            "Train Epoch: 21 [37120/60000 (62%)]\tLoss: -2056.690820\n",
            "Train Epoch: 21 [38400/60000 (64%)]\tLoss: -1970.829102\n",
            "Train Epoch: 21 [39680/60000 (66%)]\tLoss: -1986.503516\n",
            "Train Epoch: 21 [40960/60000 (68%)]\tLoss: -2019.740820\n",
            "Train Epoch: 21 [42240/60000 (70%)]\tLoss: -1975.313086\n",
            "Train Epoch: 21 [43520/60000 (72%)]\tLoss: -1933.419531\n",
            "Train Epoch: 21 [44800/60000 (75%)]\tLoss: -1993.264258\n",
            "Train Epoch: 21 [46080/60000 (77%)]\tLoss: -1973.538867\n",
            "Train Epoch: 21 [47360/60000 (79%)]\tLoss: -1963.405859\n",
            "Train Epoch: 21 [48640/60000 (81%)]\tLoss: -1959.668945\n",
            "Train Epoch: 21 [49920/60000 (83%)]\tLoss: -1984.120508\n",
            "Train Epoch: 21 [51200/60000 (85%)]\tLoss: -1999.761328\n",
            "Train Epoch: 21 [52480/60000 (87%)]\tLoss: -1943.452148\n",
            "Train Epoch: 21 [53760/60000 (90%)]\tLoss: -1970.974805\n",
            "Train Epoch: 21 [55040/60000 (92%)]\tLoss: -1941.064258\n",
            "Train Epoch: 21 [56320/60000 (94%)]\tLoss: -1963.839062\n",
            "Train Epoch: 21 [57600/60000 (96%)]\tLoss: -1949.516211\n",
            "Train Epoch: 21 [58880/60000 (98%)]\tLoss: -1931.806055\n",
            "====> Epoch: 21 Average loss: -19699.0956\n",
            "Train Epoch: 22 [0/60000 (0%)]\tLoss: -19507.025391\n",
            "Train Epoch: 22 [1280/60000 (2%)]\tLoss: -2065.233203\n",
            "Train Epoch: 22 [2560/60000 (4%)]\tLoss: -1936.856641\n",
            "Train Epoch: 22 [3840/60000 (6%)]\tLoss: -1956.299023\n",
            "Train Epoch: 22 [5120/60000 (9%)]\tLoss: -1917.167383\n",
            "Train Epoch: 22 [6400/60000 (11%)]\tLoss: -1939.613477\n",
            "Train Epoch: 22 [7680/60000 (13%)]\tLoss: -1958.082813\n",
            "Train Epoch: 22 [8960/60000 (15%)]\tLoss: -2004.907422\n",
            "Train Epoch: 22 [10240/60000 (17%)]\tLoss: -2105.225977\n",
            "Train Epoch: 22 [11520/60000 (19%)]\tLoss: -1949.537500\n",
            "Train Epoch: 22 [12800/60000 (21%)]\tLoss: -1948.538672\n",
            "Train Epoch: 22 [14080/60000 (23%)]\tLoss: -2101.971680\n",
            "Train Epoch: 22 [15360/60000 (26%)]\tLoss: -1966.178320\n",
            "Train Epoch: 22 [16640/60000 (28%)]\tLoss: -1954.030469\n",
            "Train Epoch: 22 [17920/60000 (30%)]\tLoss: -1966.129297\n",
            "Train Epoch: 22 [19200/60000 (32%)]\tLoss: -2024.356641\n",
            "Train Epoch: 22 [20480/60000 (34%)]\tLoss: -1921.128320\n",
            "Train Epoch: 22 [21760/60000 (36%)]\tLoss: -2027.964062\n",
            "Train Epoch: 22 [23040/60000 (38%)]\tLoss: -1995.929492\n",
            "Train Epoch: 22 [24320/60000 (41%)]\tLoss: -1960.056836\n",
            "Train Epoch: 22 [25600/60000 (43%)]\tLoss: -1938.241602\n",
            "Train Epoch: 22 [26880/60000 (45%)]\tLoss: -1978.452930\n",
            "Train Epoch: 22 [28160/60000 (47%)]\tLoss: -2000.699805\n",
            "Train Epoch: 22 [29440/60000 (49%)]\tLoss: -1940.995508\n",
            "Train Epoch: 22 [30720/60000 (51%)]\tLoss: -1962.934375\n",
            "Train Epoch: 22 [32000/60000 (53%)]\tLoss: -1959.061133\n",
            "Train Epoch: 22 [33280/60000 (55%)]\tLoss: -1931.245703\n",
            "Train Epoch: 22 [34560/60000 (58%)]\tLoss: -1925.314844\n",
            "Train Epoch: 22 [35840/60000 (60%)]\tLoss: -2001.084570\n",
            "Train Epoch: 22 [37120/60000 (62%)]\tLoss: -2051.592578\n",
            "Train Epoch: 22 [38400/60000 (64%)]\tLoss: -1971.109961\n",
            "Train Epoch: 22 [39680/60000 (66%)]\tLoss: -1986.120898\n",
            "Train Epoch: 22 [40960/60000 (68%)]\tLoss: -2008.961133\n",
            "Train Epoch: 22 [42240/60000 (70%)]\tLoss: -1972.713867\n",
            "Train Epoch: 22 [43520/60000 (72%)]\tLoss: -1940.717187\n",
            "Train Epoch: 22 [44800/60000 (75%)]\tLoss: -1980.420703\n",
            "Train Epoch: 22 [46080/60000 (77%)]\tLoss: -1970.825781\n",
            "Train Epoch: 22 [47360/60000 (79%)]\tLoss: -1959.138867\n",
            "Train Epoch: 22 [48640/60000 (81%)]\tLoss: -1922.920898\n",
            "Train Epoch: 22 [49920/60000 (83%)]\tLoss: -1987.783008\n",
            "Train Epoch: 22 [51200/60000 (85%)]\tLoss: -1982.954297\n",
            "Train Epoch: 22 [52480/60000 (87%)]\tLoss: -1942.383008\n",
            "Train Epoch: 22 [53760/60000 (90%)]\tLoss: -1954.362891\n",
            "Train Epoch: 22 [55040/60000 (92%)]\tLoss: -1944.604883\n",
            "Train Epoch: 22 [56320/60000 (94%)]\tLoss: -1960.738672\n",
            "Train Epoch: 22 [57600/60000 (96%)]\tLoss: -1943.657813\n",
            "Train Epoch: 22 [58880/60000 (98%)]\tLoss: -1932.613672\n",
            "====> Epoch: 22 Average loss: -19687.6562\n",
            "Train Epoch: 23 [0/60000 (0%)]\tLoss: -19479.947266\n",
            "Train Epoch: 23 [1280/60000 (2%)]\tLoss: -2076.063477\n",
            "Train Epoch: 23 [2560/60000 (4%)]\tLoss: -1954.897070\n",
            "Train Epoch: 23 [3840/60000 (6%)]\tLoss: -1956.157617\n",
            "Train Epoch: 23 [5120/60000 (9%)]\tLoss: -1954.160352\n",
            "Train Epoch: 23 [6400/60000 (11%)]\tLoss: -1941.381836\n",
            "Train Epoch: 23 [7680/60000 (13%)]\tLoss: -1955.368750\n",
            "Train Epoch: 23 [8960/60000 (15%)]\tLoss: -1978.942578\n",
            "Train Epoch: 23 [10240/60000 (17%)]\tLoss: -2097.657813\n",
            "Train Epoch: 23 [11520/60000 (19%)]\tLoss: -1948.298047\n",
            "Train Epoch: 23 [12800/60000 (21%)]\tLoss: -1942.774219\n",
            "Train Epoch: 23 [14080/60000 (23%)]\tLoss: -2142.438672\n",
            "Train Epoch: 23 [15360/60000 (26%)]\tLoss: -1959.463477\n",
            "Train Epoch: 23 [16640/60000 (28%)]\tLoss: -1956.008398\n",
            "Train Epoch: 23 [17920/60000 (30%)]\tLoss: -1946.262891\n",
            "Train Epoch: 23 [19200/60000 (32%)]\tLoss: -2001.131641\n",
            "Train Epoch: 23 [20480/60000 (34%)]\tLoss: -1921.674414\n",
            "Train Epoch: 23 [21760/60000 (36%)]\tLoss: -2041.627344\n",
            "Train Epoch: 23 [23040/60000 (38%)]\tLoss: -2017.556836\n",
            "Train Epoch: 23 [24320/60000 (41%)]\tLoss: -2005.214258\n",
            "Train Epoch: 23 [25600/60000 (43%)]\tLoss: -1917.001563\n",
            "Train Epoch: 23 [26880/60000 (45%)]\tLoss: -1967.687109\n",
            "Train Epoch: 23 [28160/60000 (47%)]\tLoss: -2003.235352\n",
            "Train Epoch: 23 [29440/60000 (49%)]\tLoss: -1955.058594\n",
            "Train Epoch: 23 [30720/60000 (51%)]\tLoss: -1965.774023\n",
            "Train Epoch: 23 [32000/60000 (53%)]\tLoss: -1941.541797\n",
            "Train Epoch: 23 [33280/60000 (55%)]\tLoss: -1933.477930\n",
            "Train Epoch: 23 [34560/60000 (58%)]\tLoss: -1922.367969\n",
            "Train Epoch: 23 [35840/60000 (60%)]\tLoss: -1966.636719\n",
            "Train Epoch: 23 [37120/60000 (62%)]\tLoss: -2034.761523\n",
            "Train Epoch: 23 [38400/60000 (64%)]\tLoss: -1968.368555\n",
            "Train Epoch: 23 [39680/60000 (66%)]\tLoss: -1980.884766\n",
            "Train Epoch: 23 [40960/60000 (68%)]\tLoss: -2017.238477\n",
            "Train Epoch: 23 [42240/60000 (70%)]\tLoss: -1969.004688\n",
            "Train Epoch: 23 [43520/60000 (72%)]\tLoss: -1943.624023\n",
            "Train Epoch: 23 [44800/60000 (75%)]\tLoss: -1989.769922\n",
            "Train Epoch: 23 [46080/60000 (77%)]\tLoss: -1968.437109\n",
            "Train Epoch: 23 [47360/60000 (79%)]\tLoss: -1978.626758\n",
            "Train Epoch: 23 [48640/60000 (81%)]\tLoss: -1952.728906\n",
            "Train Epoch: 23 [49920/60000 (83%)]\tLoss: -2015.975977\n",
            "Train Epoch: 23 [51200/60000 (85%)]\tLoss: -1970.708984\n",
            "Train Epoch: 23 [52480/60000 (87%)]\tLoss: -1954.369727\n",
            "Train Epoch: 23 [53760/60000 (90%)]\tLoss: -1962.171484\n",
            "Train Epoch: 23 [55040/60000 (92%)]\tLoss: -1942.874805\n",
            "Train Epoch: 23 [56320/60000 (94%)]\tLoss: -1962.539453\n",
            "Train Epoch: 23 [57600/60000 (96%)]\tLoss: -1941.578125\n",
            "Train Epoch: 23 [58880/60000 (98%)]\tLoss: -1928.556641\n",
            "====> Epoch: 23 Average loss: -19701.0412\n",
            "Train Epoch: 24 [0/60000 (0%)]\tLoss: -19522.064453\n",
            "Train Epoch: 24 [1280/60000 (2%)]\tLoss: -2016.912891\n",
            "Train Epoch: 24 [2560/60000 (4%)]\tLoss: -1958.261133\n",
            "Train Epoch: 24 [3840/60000 (6%)]\tLoss: -1929.844531\n",
            "Train Epoch: 24 [5120/60000 (9%)]\tLoss: -1959.504102\n",
            "Train Epoch: 24 [6400/60000 (11%)]\tLoss: -1936.526367\n",
            "Train Epoch: 24 [7680/60000 (13%)]\tLoss: -1953.102148\n",
            "Train Epoch: 24 [8960/60000 (15%)]\tLoss: -1989.016992\n",
            "Train Epoch: 24 [10240/60000 (17%)]\tLoss: -2125.988672\n",
            "Train Epoch: 24 [11520/60000 (19%)]\tLoss: -1943.624414\n",
            "Train Epoch: 24 [12800/60000 (21%)]\tLoss: -1943.259570\n",
            "Train Epoch: 24 [14080/60000 (23%)]\tLoss: -2104.393750\n",
            "Train Epoch: 24 [15360/60000 (26%)]\tLoss: -1962.478711\n",
            "Train Epoch: 24 [16640/60000 (28%)]\tLoss: -1952.955078\n",
            "Train Epoch: 24 [17920/60000 (30%)]\tLoss: -1965.683008\n",
            "Train Epoch: 24 [19200/60000 (32%)]\tLoss: -1999.476953\n",
            "Train Epoch: 24 [20480/60000 (34%)]\tLoss: -1921.489648\n",
            "Train Epoch: 24 [21760/60000 (36%)]\tLoss: -2011.885352\n",
            "Train Epoch: 24 [23040/60000 (38%)]\tLoss: -2010.947266\n",
            "Train Epoch: 24 [24320/60000 (41%)]\tLoss: -1998.900977\n",
            "Train Epoch: 24 [25600/60000 (43%)]\tLoss: -1917.576367\n",
            "Train Epoch: 24 [26880/60000 (45%)]\tLoss: -1963.191797\n",
            "Train Epoch: 24 [28160/60000 (47%)]\tLoss: -1992.340430\n",
            "Train Epoch: 24 [29440/60000 (49%)]\tLoss: -1948.330273\n",
            "Train Epoch: 24 [30720/60000 (51%)]\tLoss: -1973.923633\n",
            "Train Epoch: 24 [32000/60000 (53%)]\tLoss: -1962.829102\n",
            "Train Epoch: 24 [33280/60000 (55%)]\tLoss: -1930.668750\n",
            "Train Epoch: 24 [34560/60000 (58%)]\tLoss: -1920.020313\n",
            "Train Epoch: 24 [35840/60000 (60%)]\tLoss: -1970.786523\n",
            "Train Epoch: 24 [37120/60000 (62%)]\tLoss: -2041.757812\n",
            "Train Epoch: 24 [38400/60000 (64%)]\tLoss: -1970.184961\n",
            "Train Epoch: 24 [39680/60000 (66%)]\tLoss: -1979.383203\n",
            "Train Epoch: 24 [40960/60000 (68%)]\tLoss: -2016.700195\n",
            "Train Epoch: 24 [42240/60000 (70%)]\tLoss: -1977.416992\n",
            "Train Epoch: 24 [43520/60000 (72%)]\tLoss: -1948.115234\n",
            "Train Epoch: 24 [44800/60000 (75%)]\tLoss: -1984.862305\n",
            "Train Epoch: 24 [46080/60000 (77%)]\tLoss: -1968.072266\n",
            "Train Epoch: 24 [47360/60000 (79%)]\tLoss: -2000.313477\n",
            "Train Epoch: 24 [48640/60000 (81%)]\tLoss: -1941.908984\n",
            "Train Epoch: 24 [49920/60000 (83%)]\tLoss: -2012.248242\n",
            "Train Epoch: 24 [51200/60000 (85%)]\tLoss: -1976.751367\n",
            "Train Epoch: 24 [52480/60000 (87%)]\tLoss: -1944.773047\n",
            "Train Epoch: 24 [53760/60000 (90%)]\tLoss: -1956.684375\n",
            "Train Epoch: 24 [55040/60000 (92%)]\tLoss: -1938.018555\n",
            "Train Epoch: 24 [56320/60000 (94%)]\tLoss: -1963.991211\n",
            "Train Epoch: 24 [57600/60000 (96%)]\tLoss: -1940.620312\n",
            "Train Epoch: 24 [58880/60000 (98%)]\tLoss: -1927.962305\n",
            "====> Epoch: 24 Average loss: -19696.5235\n",
            "Train Epoch: 25 [0/60000 (0%)]\tLoss: -19667.064453\n",
            "Train Epoch: 25 [1280/60000 (2%)]\tLoss: -1997.374023\n",
            "Train Epoch: 25 [2560/60000 (4%)]\tLoss: -1955.545898\n",
            "Train Epoch: 25 [3840/60000 (6%)]\tLoss: -1952.736523\n",
            "Train Epoch: 25 [5120/60000 (9%)]\tLoss: -1976.803906\n",
            "Train Epoch: 25 [6400/60000 (11%)]\tLoss: -1938.830859\n",
            "Train Epoch: 25 [7680/60000 (13%)]\tLoss: -1948.318359\n",
            "Train Epoch: 25 [8960/60000 (15%)]\tLoss: -1971.642969\n",
            "Train Epoch: 25 [10240/60000 (17%)]\tLoss: -2170.032422\n",
            "Train Epoch: 25 [11520/60000 (19%)]\tLoss: -1941.269336\n",
            "Train Epoch: 25 [12800/60000 (21%)]\tLoss: -1945.470117\n",
            "Train Epoch: 25 [14080/60000 (23%)]\tLoss: -2074.013477\n",
            "Train Epoch: 25 [15360/60000 (26%)]\tLoss: -1961.243750\n",
            "Train Epoch: 25 [16640/60000 (28%)]\tLoss: -1952.949023\n",
            "Train Epoch: 25 [17920/60000 (30%)]\tLoss: -1957.162305\n",
            "Train Epoch: 25 [19200/60000 (32%)]\tLoss: -2005.504883\n",
            "Train Epoch: 25 [20480/60000 (34%)]\tLoss: -1919.268750\n",
            "Train Epoch: 25 [21760/60000 (36%)]\tLoss: -1982.066406\n",
            "Train Epoch: 25 [23040/60000 (38%)]\tLoss: -2020.104687\n",
            "Train Epoch: 25 [24320/60000 (41%)]\tLoss: -2006.256445\n",
            "Train Epoch: 25 [25600/60000 (43%)]\tLoss: -1919.642578\n",
            "Train Epoch: 25 [26880/60000 (45%)]\tLoss: -1962.076562\n",
            "Train Epoch: 25 [28160/60000 (47%)]\tLoss: -2026.920508\n",
            "Train Epoch: 25 [29440/60000 (49%)]\tLoss: -1953.874219\n",
            "Train Epoch: 25 [30720/60000 (51%)]\tLoss: -1973.810742\n",
            "Train Epoch: 25 [32000/60000 (53%)]\tLoss: -1960.271875\n",
            "Train Epoch: 25 [33280/60000 (55%)]\tLoss: -1929.689648\n",
            "Train Epoch: 25 [34560/60000 (58%)]\tLoss: -1935.468164\n",
            "Train Epoch: 25 [35840/60000 (60%)]\tLoss: -1988.681055\n",
            "Train Epoch: 25 [37120/60000 (62%)]\tLoss: -2021.935937\n",
            "Train Epoch: 25 [38400/60000 (64%)]\tLoss: -1986.373633\n",
            "Train Epoch: 25 [39680/60000 (66%)]\tLoss: -1960.949219\n",
            "Train Epoch: 25 [40960/60000 (68%)]\tLoss: -2010.371680\n",
            "Train Epoch: 25 [42240/60000 (70%)]\tLoss: -1968.080859\n",
            "Train Epoch: 25 [43520/60000 (72%)]\tLoss: -1945.652930\n",
            "Train Epoch: 25 [44800/60000 (75%)]\tLoss: -2001.812891\n",
            "Train Epoch: 25 [46080/60000 (77%)]\tLoss: -1968.893945\n",
            "Train Epoch: 25 [47360/60000 (79%)]\tLoss: -1953.827148\n",
            "Train Epoch: 25 [48640/60000 (81%)]\tLoss: -1954.516406\n",
            "Train Epoch: 25 [49920/60000 (83%)]\tLoss: -1984.634766\n",
            "Train Epoch: 25 [51200/60000 (85%)]\tLoss: -1941.054102\n",
            "Train Epoch: 25 [52480/60000 (87%)]\tLoss: -1944.267578\n",
            "Train Epoch: 25 [53760/60000 (90%)]\tLoss: -1949.518750\n",
            "Train Epoch: 25 [55040/60000 (92%)]\tLoss: -1927.417578\n",
            "Train Epoch: 25 [56320/60000 (94%)]\tLoss: -1961.597656\n",
            "Train Epoch: 25 [57600/60000 (96%)]\tLoss: -1943.912500\n",
            "Train Epoch: 25 [58880/60000 (98%)]\tLoss: -1923.340039\n",
            "====> Epoch: 25 Average loss: -19689.4685\n",
            "Train Epoch: 26 [0/60000 (0%)]\tLoss: -19546.304688\n",
            "Train Epoch: 26 [1280/60000 (2%)]\tLoss: -2012.333008\n",
            "Train Epoch: 26 [2560/60000 (4%)]\tLoss: -1953.226562\n",
            "Train Epoch: 26 [3840/60000 (6%)]\tLoss: -1942.700000\n",
            "Train Epoch: 26 [5120/60000 (9%)]\tLoss: -1915.883203\n",
            "Train Epoch: 26 [6400/60000 (11%)]\tLoss: -1941.816406\n",
            "Train Epoch: 26 [7680/60000 (13%)]\tLoss: -1947.286719\n",
            "Train Epoch: 26 [8960/60000 (15%)]\tLoss: -1972.120312\n",
            "Train Epoch: 26 [10240/60000 (17%)]\tLoss: -2133.998633\n",
            "Train Epoch: 26 [11520/60000 (19%)]\tLoss: -1951.164453\n",
            "Train Epoch: 26 [12800/60000 (21%)]\tLoss: -1947.153320\n",
            "Train Epoch: 26 [14080/60000 (23%)]\tLoss: -2068.582422\n",
            "Train Epoch: 26 [15360/60000 (26%)]\tLoss: -1960.176563\n",
            "Train Epoch: 26 [16640/60000 (28%)]\tLoss: -1956.394727\n",
            "Train Epoch: 26 [17920/60000 (30%)]\tLoss: -1958.405859\n",
            "Train Epoch: 26 [19200/60000 (32%)]\tLoss: -1990.666406\n",
            "Train Epoch: 26 [20480/60000 (34%)]\tLoss: -1918.696484\n",
            "Train Epoch: 26 [21760/60000 (36%)]\tLoss: -1986.722070\n",
            "Train Epoch: 26 [23040/60000 (38%)]\tLoss: -1985.999609\n",
            "Train Epoch: 26 [24320/60000 (41%)]\tLoss: -2011.507617\n",
            "Train Epoch: 26 [25600/60000 (43%)]\tLoss: -1919.683398\n",
            "Train Epoch: 26 [26880/60000 (45%)]\tLoss: -1948.568359\n",
            "Train Epoch: 26 [28160/60000 (47%)]\tLoss: -1996.970117\n",
            "Train Epoch: 26 [29440/60000 (49%)]\tLoss: -1935.350781\n",
            "Train Epoch: 26 [30720/60000 (51%)]\tLoss: -1960.950977\n",
            "Train Epoch: 26 [32000/60000 (53%)]\tLoss: -1923.207422\n",
            "Train Epoch: 26 [33280/60000 (55%)]\tLoss: -1930.365625\n",
            "Train Epoch: 26 [34560/60000 (58%)]\tLoss: -1925.651172\n",
            "Train Epoch: 26 [35840/60000 (60%)]\tLoss: -1986.488086\n",
            "Train Epoch: 26 [37120/60000 (62%)]\tLoss: -1997.329102\n",
            "Train Epoch: 26 [38400/60000 (64%)]\tLoss: -1975.628125\n",
            "Train Epoch: 26 [39680/60000 (66%)]\tLoss: -1959.245703\n",
            "Train Epoch: 26 [40960/60000 (68%)]\tLoss: -1991.212109\n",
            "Train Epoch: 26 [42240/60000 (70%)]\tLoss: -1968.411328\n",
            "Train Epoch: 26 [43520/60000 (72%)]\tLoss: -1950.666992\n",
            "Train Epoch: 26 [44800/60000 (75%)]\tLoss: -1993.826758\n",
            "Train Epoch: 26 [46080/60000 (77%)]\tLoss: -1974.140625\n",
            "Train Epoch: 26 [47360/60000 (79%)]\tLoss: -1937.335352\n",
            "Train Epoch: 26 [48640/60000 (81%)]\tLoss: -1962.756836\n",
            "Train Epoch: 26 [49920/60000 (83%)]\tLoss: -1976.331250\n",
            "Train Epoch: 26 [51200/60000 (85%)]\tLoss: -1977.339258\n",
            "Train Epoch: 26 [52480/60000 (87%)]\tLoss: -1963.489453\n",
            "Train Epoch: 26 [53760/60000 (90%)]\tLoss: -1962.905273\n",
            "Train Epoch: 26 [55040/60000 (92%)]\tLoss: -1942.600000\n",
            "Train Epoch: 26 [56320/60000 (94%)]\tLoss: -1959.162500\n",
            "Train Epoch: 26 [57600/60000 (96%)]\tLoss: -1944.154102\n",
            "Train Epoch: 26 [58880/60000 (98%)]\tLoss: -1921.875000\n",
            "====> Epoch: 26 Average loss: -19626.4678\n",
            "Train Epoch: 27 [0/60000 (0%)]\tLoss: -19523.166016\n",
            "Train Epoch: 27 [1280/60000 (2%)]\tLoss: -1981.242969\n",
            "Train Epoch: 27 [2560/60000 (4%)]\tLoss: -1938.125391\n",
            "Train Epoch: 27 [3840/60000 (6%)]\tLoss: -1941.482031\n",
            "Train Epoch: 27 [5120/60000 (9%)]\tLoss: -1910.244531\n",
            "Train Epoch: 27 [6400/60000 (11%)]\tLoss: -1941.819336\n",
            "Train Epoch: 27 [7680/60000 (13%)]\tLoss: -1967.644141\n",
            "Train Epoch: 27 [8960/60000 (15%)]\tLoss: -1974.159570\n",
            "Train Epoch: 27 [10240/60000 (17%)]\tLoss: -2119.555664\n",
            "Train Epoch: 27 [11520/60000 (19%)]\tLoss: -1950.507227\n",
            "Train Epoch: 27 [12800/60000 (21%)]\tLoss: -1942.245898\n",
            "Train Epoch: 27 [14080/60000 (23%)]\tLoss: -2053.679688\n",
            "Train Epoch: 27 [15360/60000 (26%)]\tLoss: -1959.164453\n",
            "Train Epoch: 27 [16640/60000 (28%)]\tLoss: -1951.771680\n",
            "Train Epoch: 27 [17920/60000 (30%)]\tLoss: -1961.454297\n",
            "Train Epoch: 27 [19200/60000 (32%)]\tLoss: -1983.420117\n",
            "Train Epoch: 27 [20480/60000 (34%)]\tLoss: -1918.442188\n",
            "Train Epoch: 27 [21760/60000 (36%)]\tLoss: -2018.165625\n",
            "Train Epoch: 27 [23040/60000 (38%)]\tLoss: -1967.159180\n",
            "Train Epoch: 27 [24320/60000 (41%)]\tLoss: -2005.563867\n",
            "Train Epoch: 27 [25600/60000 (43%)]\tLoss: -1920.111133\n",
            "Train Epoch: 27 [26880/60000 (45%)]\tLoss: -1942.663086\n",
            "Train Epoch: 27 [28160/60000 (47%)]\tLoss: -2016.674609\n",
            "Train Epoch: 27 [29440/60000 (49%)]\tLoss: -1939.476367\n",
            "Train Epoch: 27 [30720/60000 (51%)]\tLoss: -1963.289062\n",
            "Train Epoch: 27 [32000/60000 (53%)]\tLoss: -1934.975195\n",
            "Train Epoch: 27 [33280/60000 (55%)]\tLoss: -1927.382422\n",
            "Train Epoch: 27 [34560/60000 (58%)]\tLoss: -1925.118750\n",
            "Train Epoch: 27 [35840/60000 (60%)]\tLoss: -1975.518945\n",
            "Train Epoch: 27 [37120/60000 (62%)]\tLoss: -2017.149609\n",
            "Train Epoch: 27 [38400/60000 (64%)]\tLoss: -1971.238672\n",
            "Train Epoch: 27 [39680/60000 (66%)]\tLoss: -1959.180859\n",
            "Train Epoch: 27 [40960/60000 (68%)]\tLoss: -2026.912305\n",
            "Train Epoch: 27 [42240/60000 (70%)]\tLoss: -1965.632812\n",
            "Train Epoch: 27 [43520/60000 (72%)]\tLoss: -1947.935156\n",
            "Train Epoch: 27 [44800/60000 (75%)]\tLoss: -1996.261133\n",
            "Train Epoch: 27 [46080/60000 (77%)]\tLoss: -1967.924414\n",
            "Train Epoch: 27 [47360/60000 (79%)]\tLoss: -1950.121680\n",
            "Train Epoch: 27 [48640/60000 (81%)]\tLoss: -1971.588281\n",
            "Train Epoch: 27 [49920/60000 (83%)]\tLoss: -1986.756055\n",
            "Train Epoch: 27 [51200/60000 (85%)]\tLoss: -1935.024609\n",
            "Train Epoch: 27 [52480/60000 (87%)]\tLoss: -1960.441211\n",
            "Train Epoch: 27 [53760/60000 (90%)]\tLoss: -1948.890039\n",
            "Train Epoch: 27 [55040/60000 (92%)]\tLoss: -1939.741797\n",
            "Train Epoch: 27 [56320/60000 (94%)]\tLoss: -1958.427148\n",
            "Train Epoch: 27 [57600/60000 (96%)]\tLoss: -1952.192383\n",
            "Train Epoch: 27 [58880/60000 (98%)]\tLoss: -1923.825977\n",
            "====> Epoch: 27 Average loss: -19641.1228\n",
            "Train Epoch: 28 [0/60000 (0%)]\tLoss: -19559.033203\n",
            "Train Epoch: 28 [1280/60000 (2%)]\tLoss: -2019.462500\n",
            "Train Epoch: 28 [2560/60000 (4%)]\tLoss: -1949.903125\n",
            "Train Epoch: 28 [3840/60000 (6%)]\tLoss: -1944.327734\n",
            "Train Epoch: 28 [5120/60000 (9%)]\tLoss: -1917.520117\n",
            "Train Epoch: 28 [6400/60000 (11%)]\tLoss: -1947.135352\n",
            "Train Epoch: 28 [7680/60000 (13%)]\tLoss: -1949.890820\n",
            "Train Epoch: 28 [8960/60000 (15%)]\tLoss: -1970.711914\n",
            "Train Epoch: 28 [10240/60000 (17%)]\tLoss: -2170.835938\n",
            "Train Epoch: 28 [11520/60000 (19%)]\tLoss: -1960.086328\n",
            "Train Epoch: 28 [12800/60000 (21%)]\tLoss: -1956.420117\n",
            "Train Epoch: 28 [14080/60000 (23%)]\tLoss: -2065.190234\n",
            "Train Epoch: 28 [15360/60000 (26%)]\tLoss: -1961.437305\n",
            "Train Epoch: 28 [16640/60000 (28%)]\tLoss: -1951.113477\n",
            "Train Epoch: 28 [17920/60000 (30%)]\tLoss: -1961.755078\n",
            "Train Epoch: 28 [19200/60000 (32%)]\tLoss: -1983.165234\n",
            "Train Epoch: 28 [20480/60000 (34%)]\tLoss: -1917.278516\n",
            "Train Epoch: 28 [21760/60000 (36%)]\tLoss: -2031.800195\n",
            "Train Epoch: 28 [23040/60000 (38%)]\tLoss: -1975.402148\n",
            "Train Epoch: 28 [24320/60000 (41%)]\tLoss: -2010.868555\n",
            "Train Epoch: 28 [25600/60000 (43%)]\tLoss: -1919.266602\n",
            "Train Epoch: 28 [26880/60000 (45%)]\tLoss: -1945.080664\n",
            "Train Epoch: 28 [28160/60000 (47%)]\tLoss: -2044.419141\n",
            "Train Epoch: 28 [29440/60000 (49%)]\tLoss: -1956.675391\n",
            "Train Epoch: 28 [30720/60000 (51%)]\tLoss: -1974.528711\n",
            "Train Epoch: 28 [32000/60000 (53%)]\tLoss: -1932.564258\n",
            "Train Epoch: 28 [33280/60000 (55%)]\tLoss: -1926.220313\n",
            "Train Epoch: 28 [34560/60000 (58%)]\tLoss: -1921.995312\n",
            "Train Epoch: 28 [35840/60000 (60%)]\tLoss: -1980.427148\n",
            "Train Epoch: 28 [37120/60000 (62%)]\tLoss: -2025.404297\n",
            "Train Epoch: 28 [38400/60000 (64%)]\tLoss: -1966.377539\n",
            "Train Epoch: 28 [39680/60000 (66%)]\tLoss: -1958.994922\n",
            "Train Epoch: 28 [40960/60000 (68%)]\tLoss: -2021.490625\n",
            "Train Epoch: 28 [42240/60000 (70%)]\tLoss: -1981.866016\n",
            "Train Epoch: 28 [43520/60000 (72%)]\tLoss: -1953.409570\n",
            "Train Epoch: 28 [44800/60000 (75%)]\tLoss: -2012.986328\n",
            "Train Epoch: 28 [46080/60000 (77%)]\tLoss: -1967.764648\n",
            "Train Epoch: 28 [47360/60000 (79%)]\tLoss: -1967.341211\n",
            "Train Epoch: 28 [48640/60000 (81%)]\tLoss: -1962.208008\n",
            "Train Epoch: 28 [49920/60000 (83%)]\tLoss: -1986.298633\n",
            "Train Epoch: 28 [51200/60000 (85%)]\tLoss: -1958.290430\n",
            "Train Epoch: 28 [52480/60000 (87%)]\tLoss: -1982.968164\n",
            "Train Epoch: 28 [53760/60000 (90%)]\tLoss: -1962.749805\n",
            "Train Epoch: 28 [55040/60000 (92%)]\tLoss: -1922.121289\n",
            "Train Epoch: 28 [56320/60000 (94%)]\tLoss: -1960.657031\n",
            "Train Epoch: 28 [57600/60000 (96%)]\tLoss: -1959.208008\n",
            "Train Epoch: 28 [58880/60000 (98%)]\tLoss: -1921.943164\n",
            "====> Epoch: 28 Average loss: -19671.2617\n",
            "Train Epoch: 29 [0/60000 (0%)]\tLoss: -19401.294922\n",
            "Train Epoch: 29 [1280/60000 (2%)]\tLoss: -2055.064648\n",
            "Train Epoch: 29 [2560/60000 (4%)]\tLoss: -1950.312891\n",
            "Train Epoch: 29 [3840/60000 (6%)]\tLoss: -1942.791406\n",
            "Train Epoch: 29 [5120/60000 (9%)]\tLoss: -1926.903125\n",
            "Train Epoch: 29 [6400/60000 (11%)]\tLoss: -1941.806250\n",
            "Train Epoch: 29 [7680/60000 (13%)]\tLoss: -1948.332227\n",
            "Train Epoch: 29 [8960/60000 (15%)]\tLoss: -1970.720117\n",
            "Train Epoch: 29 [10240/60000 (17%)]\tLoss: -2149.753516\n",
            "Train Epoch: 29 [11520/60000 (19%)]\tLoss: -1946.458398\n",
            "Train Epoch: 29 [12800/60000 (21%)]\tLoss: -1973.253125\n",
            "Train Epoch: 29 [14080/60000 (23%)]\tLoss: -2126.432813\n",
            "Train Epoch: 29 [15360/60000 (26%)]\tLoss: -1959.949023\n",
            "Train Epoch: 29 [16640/60000 (28%)]\tLoss: -1951.429297\n",
            "Train Epoch: 29 [17920/60000 (30%)]\tLoss: -1960.515430\n",
            "Train Epoch: 29 [19200/60000 (32%)]\tLoss: -1993.465234\n",
            "Train Epoch: 29 [20480/60000 (34%)]\tLoss: -1921.293164\n",
            "Train Epoch: 29 [21760/60000 (36%)]\tLoss: -1993.575781\n",
            "Train Epoch: 29 [23040/60000 (38%)]\tLoss: -1989.581055\n",
            "Train Epoch: 29 [24320/60000 (41%)]\tLoss: -2038.688867\n",
            "Train Epoch: 29 [25600/60000 (43%)]\tLoss: -1918.269336\n",
            "Train Epoch: 29 [26880/60000 (45%)]\tLoss: -1958.115625\n",
            "Train Epoch: 29 [28160/60000 (47%)]\tLoss: -2013.941602\n",
            "Train Epoch: 29 [29440/60000 (49%)]\tLoss: -1959.070117\n",
            "Train Epoch: 29 [30720/60000 (51%)]\tLoss: -1970.568945\n",
            "Train Epoch: 29 [32000/60000 (53%)]\tLoss: -1941.580859\n",
            "Train Epoch: 29 [33280/60000 (55%)]\tLoss: -1925.060156\n",
            "Train Epoch: 29 [34560/60000 (58%)]\tLoss: -1921.272461\n",
            "Train Epoch: 29 [35840/60000 (60%)]\tLoss: -1982.469336\n",
            "Train Epoch: 29 [37120/60000 (62%)]\tLoss: -2016.961523\n",
            "Train Epoch: 29 [38400/60000 (64%)]\tLoss: -1968.995312\n",
            "Train Epoch: 29 [39680/60000 (66%)]\tLoss: -1958.276953\n",
            "Train Epoch: 29 [40960/60000 (68%)]\tLoss: -1998.933789\n",
            "Train Epoch: 29 [42240/60000 (70%)]\tLoss: -1971.934180\n",
            "Train Epoch: 29 [43520/60000 (72%)]\tLoss: -1946.787695\n",
            "Train Epoch: 29 [44800/60000 (75%)]\tLoss: -1979.842383\n",
            "Train Epoch: 29 [46080/60000 (77%)]\tLoss: -2004.220898\n",
            "Train Epoch: 29 [47360/60000 (79%)]\tLoss: -1949.138281\n",
            "Train Epoch: 29 [48640/60000 (81%)]\tLoss: -1944.440234\n",
            "Train Epoch: 29 [49920/60000 (83%)]\tLoss: -1981.619531\n",
            "Train Epoch: 29 [51200/60000 (85%)]\tLoss: -1968.551758\n",
            "Train Epoch: 29 [52480/60000 (87%)]\tLoss: -1959.233008\n",
            "Train Epoch: 29 [53760/60000 (90%)]\tLoss: -1964.347070\n",
            "Train Epoch: 29 [55040/60000 (92%)]\tLoss: -1939.211914\n",
            "Train Epoch: 29 [56320/60000 (94%)]\tLoss: -1978.797266\n",
            "Train Epoch: 29 [57600/60000 (96%)]\tLoss: -1962.890039\n",
            "Train Epoch: 29 [58880/60000 (98%)]\tLoss: -1916.088281\n",
            "====> Epoch: 29 Average loss: -19688.3255\n",
            "Train Epoch: 30 [0/60000 (0%)]\tLoss: -19332.996094\n",
            "Train Epoch: 30 [1280/60000 (2%)]\tLoss: -2072.316602\n",
            "Train Epoch: 30 [2560/60000 (4%)]\tLoss: -1948.623633\n",
            "Train Epoch: 30 [3840/60000 (6%)]\tLoss: -1944.272656\n",
            "Train Epoch: 30 [5120/60000 (9%)]\tLoss: -1930.846289\n",
            "Train Epoch: 30 [6400/60000 (11%)]\tLoss: -1938.577734\n",
            "Train Epoch: 30 [7680/60000 (13%)]\tLoss: -1948.447656\n",
            "Train Epoch: 30 [8960/60000 (15%)]\tLoss: -1968.528906\n",
            "Train Epoch: 30 [10240/60000 (17%)]\tLoss: -2139.647070\n",
            "Train Epoch: 30 [11520/60000 (19%)]\tLoss: -1962.383203\n",
            "Train Epoch: 30 [12800/60000 (21%)]\tLoss: -1963.789648\n",
            "Train Epoch: 30 [14080/60000 (23%)]\tLoss: -2072.774023\n",
            "Train Epoch: 30 [15360/60000 (26%)]\tLoss: -1959.473828\n",
            "Train Epoch: 30 [16640/60000 (28%)]\tLoss: -1949.730469\n",
            "Train Epoch: 30 [17920/60000 (30%)]\tLoss: -1956.643555\n",
            "Train Epoch: 30 [19200/60000 (32%)]\tLoss: -1994.368945\n",
            "Train Epoch: 30 [20480/60000 (34%)]\tLoss: -1915.847656\n",
            "Train Epoch: 30 [21760/60000 (36%)]\tLoss: -2011.901953\n",
            "Train Epoch: 30 [23040/60000 (38%)]\tLoss: -1978.711523\n",
            "Train Epoch: 30 [24320/60000 (41%)]\tLoss: -2022.601953\n",
            "Train Epoch: 30 [25600/60000 (43%)]\tLoss: -1919.357031\n",
            "Train Epoch: 30 [26880/60000 (45%)]\tLoss: -1952.196484\n",
            "Train Epoch: 30 [28160/60000 (47%)]\tLoss: -2041.929492\n",
            "Train Epoch: 30 [29440/60000 (49%)]\tLoss: -1955.150781\n",
            "Train Epoch: 30 [30720/60000 (51%)]\tLoss: -1960.308789\n",
            "Train Epoch: 30 [32000/60000 (53%)]\tLoss: -1941.547461\n",
            "Train Epoch: 30 [33280/60000 (55%)]\tLoss: -1923.718750\n",
            "Train Epoch: 30 [34560/60000 (58%)]\tLoss: -1917.824609\n",
            "Train Epoch: 30 [35840/60000 (60%)]\tLoss: -1973.818555\n",
            "Train Epoch: 30 [37120/60000 (62%)]\tLoss: -2045.949805\n",
            "Train Epoch: 30 [38400/60000 (64%)]\tLoss: -1974.207617\n",
            "Train Epoch: 30 [39680/60000 (66%)]\tLoss: -1958.296875\n",
            "Train Epoch: 30 [40960/60000 (68%)]\tLoss: -2020.767188\n",
            "Train Epoch: 30 [42240/60000 (70%)]\tLoss: -1951.144336\n",
            "Train Epoch: 30 [43520/60000 (72%)]\tLoss: -1946.875195\n",
            "Train Epoch: 30 [44800/60000 (75%)]\tLoss: -1983.292578\n",
            "Train Epoch: 30 [46080/60000 (77%)]\tLoss: -1987.207031\n",
            "Train Epoch: 30 [47360/60000 (79%)]\tLoss: -1972.247852\n",
            "Train Epoch: 30 [48640/60000 (81%)]\tLoss: -1970.528516\n",
            "Train Epoch: 30 [49920/60000 (83%)]\tLoss: -1973.999219\n",
            "Train Epoch: 30 [51200/60000 (85%)]\tLoss: -1946.747852\n",
            "Train Epoch: 30 [52480/60000 (87%)]\tLoss: -1964.708398\n",
            "Train Epoch: 30 [53760/60000 (90%)]\tLoss: -1960.626367\n",
            "Train Epoch: 30 [55040/60000 (92%)]\tLoss: -1939.224219\n",
            "Train Epoch: 30 [56320/60000 (94%)]\tLoss: -1965.733789\n",
            "Train Epoch: 30 [57600/60000 (96%)]\tLoss: -1963.730469\n",
            "Train Epoch: 30 [58880/60000 (98%)]\tLoss: -1916.982227\n",
            "====> Epoch: 30 Average loss: -19686.4222\n",
            "Train Epoch: 31 [0/60000 (0%)]\tLoss: -19524.941406\n",
            "Train Epoch: 31 [1280/60000 (2%)]\tLoss: -2043.379688\n",
            "Train Epoch: 31 [2560/60000 (4%)]\tLoss: -1947.418945\n",
            "Train Epoch: 31 [3840/60000 (6%)]\tLoss: -1959.564063\n",
            "Train Epoch: 31 [5120/60000 (9%)]\tLoss: -1921.919922\n",
            "Train Epoch: 31 [6400/60000 (11%)]\tLoss: -1940.616992\n",
            "Train Epoch: 31 [7680/60000 (13%)]\tLoss: -1953.300195\n",
            "Train Epoch: 31 [8960/60000 (15%)]\tLoss: -1969.223242\n",
            "Train Epoch: 31 [10240/60000 (17%)]\tLoss: -2196.458984\n",
            "Train Epoch: 31 [11520/60000 (19%)]\tLoss: -1961.904492\n",
            "Train Epoch: 31 [12800/60000 (21%)]\tLoss: -1943.144141\n",
            "Train Epoch: 31 [14080/60000 (23%)]\tLoss: -2117.250391\n",
            "Train Epoch: 31 [15360/60000 (26%)]\tLoss: -1959.234570\n",
            "Train Epoch: 31 [16640/60000 (28%)]\tLoss: -1950.592187\n",
            "Train Epoch: 31 [17920/60000 (30%)]\tLoss: -1958.967773\n",
            "Train Epoch: 31 [19200/60000 (32%)]\tLoss: -1989.754102\n",
            "Train Epoch: 31 [20480/60000 (34%)]\tLoss: -1914.501563\n",
            "Train Epoch: 31 [21760/60000 (36%)]\tLoss: -2032.886914\n",
            "Train Epoch: 31 [23040/60000 (38%)]\tLoss: -1982.700781\n",
            "Train Epoch: 31 [24320/60000 (41%)]\tLoss: -2023.198437\n",
            "Train Epoch: 31 [25600/60000 (43%)]\tLoss: -1917.410547\n",
            "Train Epoch: 31 [26880/60000 (45%)]\tLoss: -1955.754883\n",
            "Train Epoch: 31 [28160/60000 (47%)]\tLoss: -2040.375977\n",
            "Train Epoch: 31 [29440/60000 (49%)]\tLoss: -1936.066602\n",
            "Train Epoch: 31 [30720/60000 (51%)]\tLoss: -1970.624414\n",
            "Train Epoch: 31 [32000/60000 (53%)]\tLoss: -1955.452148\n",
            "Train Epoch: 31 [33280/60000 (55%)]\tLoss: -1925.226953\n",
            "Train Epoch: 31 [34560/60000 (58%)]\tLoss: -1937.198437\n",
            "Train Epoch: 31 [35840/60000 (60%)]\tLoss: -1992.312305\n",
            "Train Epoch: 31 [37120/60000 (62%)]\tLoss: -2021.788086\n",
            "Train Epoch: 31 [38400/60000 (64%)]\tLoss: -1965.382617\n",
            "Train Epoch: 31 [39680/60000 (66%)]\tLoss: -1956.463672\n",
            "Train Epoch: 31 [40960/60000 (68%)]\tLoss: -2036.603516\n",
            "Train Epoch: 31 [42240/60000 (70%)]\tLoss: -2001.922070\n",
            "Train Epoch: 31 [43520/60000 (72%)]\tLoss: -1947.838086\n",
            "Train Epoch: 31 [44800/60000 (75%)]\tLoss: -1983.797070\n",
            "Train Epoch: 31 [46080/60000 (77%)]\tLoss: -1963.166211\n",
            "Train Epoch: 31 [47360/60000 (79%)]\tLoss: -1964.297852\n",
            "Train Epoch: 31 [48640/60000 (81%)]\tLoss: -1982.295117\n",
            "Train Epoch: 31 [49920/60000 (83%)]\tLoss: -1985.506055\n",
            "Train Epoch: 31 [51200/60000 (85%)]\tLoss: -1966.062695\n",
            "Train Epoch: 31 [52480/60000 (87%)]\tLoss: -1976.532813\n",
            "Train Epoch: 31 [53760/60000 (90%)]\tLoss: -1962.703906\n",
            "Train Epoch: 31 [55040/60000 (92%)]\tLoss: -1937.841211\n",
            "Train Epoch: 31 [56320/60000 (94%)]\tLoss: -1968.690625\n",
            "Train Epoch: 31 [57600/60000 (96%)]\tLoss: -1964.681641\n",
            "Train Epoch: 31 [58880/60000 (98%)]\tLoss: -1918.784766\n",
            "====> Epoch: 31 Average loss: -19714.1050\n",
            "Train Epoch: 32 [0/60000 (0%)]\tLoss: -19282.148438\n",
            "Train Epoch: 32 [1280/60000 (2%)]\tLoss: -2077.578906\n",
            "Train Epoch: 32 [2560/60000 (4%)]\tLoss: -1945.324805\n",
            "Train Epoch: 32 [3840/60000 (6%)]\tLoss: -1964.809766\n",
            "Train Epoch: 32 [5120/60000 (9%)]\tLoss: -1943.298047\n",
            "Train Epoch: 32 [6400/60000 (11%)]\tLoss: -1939.027539\n",
            "Train Epoch: 32 [7680/60000 (13%)]\tLoss: -1953.762695\n",
            "Train Epoch: 32 [8960/60000 (15%)]\tLoss: -1990.361133\n",
            "Train Epoch: 32 [10240/60000 (17%)]\tLoss: -2153.496289\n",
            "Train Epoch: 32 [11520/60000 (19%)]\tLoss: -1961.908008\n",
            "Train Epoch: 32 [12800/60000 (21%)]\tLoss: -1941.263086\n",
            "Train Epoch: 32 [14080/60000 (23%)]\tLoss: -2162.691797\n",
            "Train Epoch: 32 [15360/60000 (26%)]\tLoss: -1958.718164\n",
            "Train Epoch: 32 [16640/60000 (28%)]\tLoss: -1950.336523\n",
            "Train Epoch: 32 [17920/60000 (30%)]\tLoss: -1959.098633\n",
            "Train Epoch: 32 [19200/60000 (32%)]\tLoss: -2016.171289\n",
            "Train Epoch: 32 [20480/60000 (34%)]\tLoss: -1915.810547\n",
            "Train Epoch: 32 [21760/60000 (36%)]\tLoss: -2017.776758\n",
            "Train Epoch: 32 [23040/60000 (38%)]\tLoss: -2026.080859\n",
            "Train Epoch: 32 [24320/60000 (41%)]\tLoss: -2009.054883\n",
            "Train Epoch: 32 [25600/60000 (43%)]\tLoss: -1918.329102\n",
            "Train Epoch: 32 [26880/60000 (45%)]\tLoss: -1949.130859\n",
            "Train Epoch: 32 [28160/60000 (47%)]\tLoss: -2037.029102\n",
            "Train Epoch: 32 [29440/60000 (49%)]\tLoss: -1959.878906\n",
            "Train Epoch: 32 [30720/60000 (51%)]\tLoss: -1966.395703\n",
            "Train Epoch: 32 [32000/60000 (53%)]\tLoss: -1944.266211\n",
            "Train Epoch: 32 [33280/60000 (55%)]\tLoss: -1930.831250\n",
            "Train Epoch: 32 [34560/60000 (58%)]\tLoss: -1944.780078\n",
            "Train Epoch: 32 [35840/60000 (60%)]\tLoss: -1979.612500\n",
            "Train Epoch: 32 [37120/60000 (62%)]\tLoss: -2000.278906\n",
            "Train Epoch: 32 [38400/60000 (64%)]\tLoss: -1966.674805\n",
            "Train Epoch: 32 [39680/60000 (66%)]\tLoss: -1952.309180\n",
            "Train Epoch: 32 [40960/60000 (68%)]\tLoss: -2041.632227\n",
            "Train Epoch: 32 [42240/60000 (70%)]\tLoss: -1958.781836\n",
            "Train Epoch: 32 [43520/60000 (72%)]\tLoss: -1951.471094\n",
            "Train Epoch: 32 [44800/60000 (75%)]\tLoss: -1987.916992\n",
            "Train Epoch: 32 [46080/60000 (77%)]\tLoss: -1963.860938\n",
            "Train Epoch: 32 [47360/60000 (79%)]\tLoss: -1968.517188\n",
            "Train Epoch: 32 [48640/60000 (81%)]\tLoss: -1951.517773\n",
            "Train Epoch: 32 [49920/60000 (83%)]\tLoss: -1989.334375\n",
            "Train Epoch: 32 [51200/60000 (85%)]\tLoss: -1944.688477\n",
            "Train Epoch: 32 [52480/60000 (87%)]\tLoss: -1952.731445\n",
            "Train Epoch: 32 [53760/60000 (90%)]\tLoss: -1959.663477\n",
            "Train Epoch: 32 [55040/60000 (92%)]\tLoss: -1919.729102\n",
            "Train Epoch: 32 [56320/60000 (94%)]\tLoss: -1957.268359\n",
            "Train Epoch: 32 [57600/60000 (96%)]\tLoss: -1961.317578\n",
            "Train Epoch: 32 [58880/60000 (98%)]\tLoss: -1916.783594\n",
            "====> Epoch: 32 Average loss: -19696.0717\n",
            "Train Epoch: 33 [0/60000 (0%)]\tLoss: -19284.720703\n",
            "Train Epoch: 33 [1280/60000 (2%)]\tLoss: -2054.583594\n",
            "Train Epoch: 33 [2560/60000 (4%)]\tLoss: -1922.028906\n",
            "Train Epoch: 33 [3840/60000 (6%)]\tLoss: -1959.298047\n",
            "Train Epoch: 33 [5120/60000 (9%)]\tLoss: -1923.745898\n",
            "Train Epoch: 33 [6400/60000 (11%)]\tLoss: -1938.541211\n",
            "Train Epoch: 33 [7680/60000 (13%)]\tLoss: -1953.852930\n",
            "Train Epoch: 33 [8960/60000 (15%)]\tLoss: -1978.766992\n",
            "Train Epoch: 33 [10240/60000 (17%)]\tLoss: -2164.318164\n",
            "Train Epoch: 33 [11520/60000 (19%)]\tLoss: -1953.219336\n",
            "Train Epoch: 33 [12800/60000 (21%)]\tLoss: -1941.708203\n",
            "Train Epoch: 33 [14080/60000 (23%)]\tLoss: -2067.893555\n",
            "Train Epoch: 33 [15360/60000 (26%)]\tLoss: -1957.687109\n",
            "Train Epoch: 33 [16640/60000 (28%)]\tLoss: -1953.375977\n",
            "Train Epoch: 33 [17920/60000 (30%)]\tLoss: -1956.157813\n",
            "Train Epoch: 33 [19200/60000 (32%)]\tLoss: -1981.937500\n",
            "Train Epoch: 33 [20480/60000 (34%)]\tLoss: -1913.465430\n",
            "Train Epoch: 33 [21760/60000 (36%)]\tLoss: -2003.841016\n",
            "Train Epoch: 33 [23040/60000 (38%)]\tLoss: -1987.218164\n",
            "Train Epoch: 33 [24320/60000 (41%)]\tLoss: -1986.746875\n",
            "Train Epoch: 33 [25600/60000 (43%)]\tLoss: -1918.162305\n",
            "Train Epoch: 33 [26880/60000 (45%)]\tLoss: -1952.107227\n",
            "Train Epoch: 33 [28160/60000 (47%)]\tLoss: -2011.605859\n",
            "Train Epoch: 33 [29440/60000 (49%)]\tLoss: -1942.986523\n",
            "Train Epoch: 33 [30720/60000 (51%)]\tLoss: -1954.450977\n",
            "Train Epoch: 33 [32000/60000 (53%)]\tLoss: -1942.764062\n",
            "Train Epoch: 33 [33280/60000 (55%)]\tLoss: -1924.230469\n",
            "Train Epoch: 33 [34560/60000 (58%)]\tLoss: -1919.437305\n",
            "Train Epoch: 33 [35840/60000 (60%)]\tLoss: -1976.732031\n",
            "Train Epoch: 33 [37120/60000 (62%)]\tLoss: -2010.248828\n",
            "Train Epoch: 33 [38400/60000 (64%)]\tLoss: -1964.575781\n",
            "Train Epoch: 33 [39680/60000 (66%)]\tLoss: -1954.928320\n",
            "Train Epoch: 33 [40960/60000 (68%)]\tLoss: -1987.525391\n",
            "Train Epoch: 33 [42240/60000 (70%)]\tLoss: -1975.216602\n",
            "Train Epoch: 33 [43520/60000 (72%)]\tLoss: -1942.797461\n",
            "Train Epoch: 33 [44800/60000 (75%)]\tLoss: -1980.969531\n",
            "Train Epoch: 33 [46080/60000 (77%)]\tLoss: -1984.973047\n",
            "Train Epoch: 33 [47360/60000 (79%)]\tLoss: -1975.150977\n",
            "Train Epoch: 33 [48640/60000 (81%)]\tLoss: -1973.546484\n",
            "Train Epoch: 33 [49920/60000 (83%)]\tLoss: -1997.342187\n",
            "Train Epoch: 33 [51200/60000 (85%)]\tLoss: -1927.435352\n",
            "Train Epoch: 33 [52480/60000 (87%)]\tLoss: -1973.542578\n",
            "Train Epoch: 33 [53760/60000 (90%)]\tLoss: -1959.307617\n",
            "Train Epoch: 33 [55040/60000 (92%)]\tLoss: -1935.716992\n",
            "Train Epoch: 33 [56320/60000 (94%)]\tLoss: -1963.365625\n",
            "Train Epoch: 33 [57600/60000 (96%)]\tLoss: -1961.606641\n",
            "Train Epoch: 33 [58880/60000 (98%)]\tLoss: -1910.351172\n",
            "====> Epoch: 33 Average loss: -19633.6437\n",
            "Train Epoch: 34 [0/60000 (0%)]\tLoss: -19229.777344\n",
            "Train Epoch: 34 [1280/60000 (2%)]\tLoss: -1999.724023\n",
            "Train Epoch: 34 [2560/60000 (4%)]\tLoss: -1924.383789\n",
            "Train Epoch: 34 [3840/60000 (6%)]\tLoss: -1950.905273\n",
            "Train Epoch: 34 [5120/60000 (9%)]\tLoss: -1967.842969\n",
            "Train Epoch: 34 [6400/60000 (11%)]\tLoss: -1929.033008\n",
            "Train Epoch: 34 [7680/60000 (13%)]\tLoss: -1952.135547\n",
            "Train Epoch: 34 [8960/60000 (15%)]\tLoss: -1973.939258\n",
            "Train Epoch: 34 [10240/60000 (17%)]\tLoss: -2158.107031\n",
            "Train Epoch: 34 [11520/60000 (19%)]\tLoss: -1943.939258\n",
            "Train Epoch: 34 [12800/60000 (21%)]\tLoss: -1948.699609\n",
            "Train Epoch: 34 [14080/60000 (23%)]\tLoss: -2073.771680\n",
            "Train Epoch: 34 [15360/60000 (26%)]\tLoss: -1958.629102\n",
            "Train Epoch: 34 [16640/60000 (28%)]\tLoss: -1951.074023\n",
            "Train Epoch: 34 [17920/60000 (30%)]\tLoss: -1956.205078\n",
            "Train Epoch: 34 [19200/60000 (32%)]\tLoss: -1977.597852\n",
            "Train Epoch: 34 [20480/60000 (34%)]\tLoss: -1911.408984\n",
            "Train Epoch: 34 [21760/60000 (36%)]\tLoss: -1980.366211\n",
            "Train Epoch: 34 [23040/60000 (38%)]\tLoss: -1965.393164\n",
            "Train Epoch: 34 [24320/60000 (41%)]\tLoss: -2049.231836\n",
            "Train Epoch: 34 [25600/60000 (43%)]\tLoss: -1914.394336\n",
            "Train Epoch: 34 [26880/60000 (45%)]\tLoss: -1956.711914\n",
            "Train Epoch: 34 [28160/60000 (47%)]\tLoss: -1977.185937\n",
            "Train Epoch: 34 [29440/60000 (49%)]\tLoss: -1937.212891\n",
            "Train Epoch: 34 [30720/60000 (51%)]\tLoss: -1963.057617\n",
            "Train Epoch: 34 [32000/60000 (53%)]\tLoss: -1936.780078\n",
            "Train Epoch: 34 [33280/60000 (55%)]\tLoss: -1929.113672\n",
            "Train Epoch: 34 [34560/60000 (58%)]\tLoss: -1921.241211\n",
            "Train Epoch: 34 [35840/60000 (60%)]\tLoss: -1986.882422\n",
            "Train Epoch: 34 [37120/60000 (62%)]\tLoss: -2007.941016\n",
            "Train Epoch: 34 [38400/60000 (64%)]\tLoss: -1965.282422\n",
            "Train Epoch: 34 [39680/60000 (66%)]\tLoss: -1958.548047\n",
            "Train Epoch: 34 [40960/60000 (68%)]\tLoss: -2003.065039\n",
            "Train Epoch: 34 [42240/60000 (70%)]\tLoss: -1958.014844\n",
            "Train Epoch: 34 [43520/60000 (72%)]\tLoss: -1940.976758\n",
            "Train Epoch: 34 [44800/60000 (75%)]\tLoss: -2007.509766\n",
            "Train Epoch: 34 [46080/60000 (77%)]\tLoss: -1968.450977\n",
            "Train Epoch: 34 [47360/60000 (79%)]\tLoss: -1977.217187\n",
            "Train Epoch: 34 [48640/60000 (81%)]\tLoss: -1956.513281\n",
            "Train Epoch: 34 [49920/60000 (83%)]\tLoss: -2003.752344\n",
            "Train Epoch: 34 [51200/60000 (85%)]\tLoss: -1994.818359\n",
            "Train Epoch: 34 [52480/60000 (87%)]\tLoss: -1972.485352\n",
            "Train Epoch: 34 [53760/60000 (90%)]\tLoss: -1958.627148\n",
            "Train Epoch: 34 [55040/60000 (92%)]\tLoss: -1931.822070\n",
            "Train Epoch: 34 [56320/60000 (94%)]\tLoss: -1956.762695\n",
            "Train Epoch: 34 [57600/60000 (96%)]\tLoss: -1958.689844\n",
            "Train Epoch: 34 [58880/60000 (98%)]\tLoss: -1918.400391\n",
            "====> Epoch: 34 Average loss: -19656.5209\n",
            "Train Epoch: 35 [0/60000 (0%)]\tLoss: -19340.328125\n",
            "Train Epoch: 35 [1280/60000 (2%)]\tLoss: -2012.002930\n",
            "Train Epoch: 35 [2560/60000 (4%)]\tLoss: -1933.930078\n",
            "Train Epoch: 35 [3840/60000 (6%)]\tLoss: -1956.930078\n",
            "Train Epoch: 35 [5120/60000 (9%)]\tLoss: -1951.226953\n",
            "Train Epoch: 35 [6400/60000 (11%)]\tLoss: -1932.142578\n",
            "Train Epoch: 35 [7680/60000 (13%)]\tLoss: -1948.936719\n",
            "Train Epoch: 35 [8960/60000 (15%)]\tLoss: -1978.133203\n",
            "Train Epoch: 35 [10240/60000 (17%)]\tLoss: -2191.533203\n",
            "Train Epoch: 35 [11520/60000 (19%)]\tLoss: -1963.588867\n",
            "Train Epoch: 35 [12800/60000 (21%)]\tLoss: -1948.492969\n",
            "Train Epoch: 35 [14080/60000 (23%)]\tLoss: -2105.852930\n",
            "Train Epoch: 35 [15360/60000 (26%)]\tLoss: -1956.259961\n",
            "Train Epoch: 35 [16640/60000 (28%)]\tLoss: -1954.618164\n",
            "Train Epoch: 35 [17920/60000 (30%)]\tLoss: -1952.470703\n",
            "Train Epoch: 35 [19200/60000 (32%)]\tLoss: -2011.959570\n",
            "Train Epoch: 35 [20480/60000 (34%)]\tLoss: -1907.584180\n",
            "Train Epoch: 35 [21760/60000 (36%)]\tLoss: -2018.791602\n",
            "Train Epoch: 35 [23040/60000 (38%)]\tLoss: -1974.827148\n",
            "Train Epoch: 35 [24320/60000 (41%)]\tLoss: -2045.615430\n",
            "Train Epoch: 35 [25600/60000 (43%)]\tLoss: -1923.781836\n",
            "Train Epoch: 35 [26880/60000 (45%)]\tLoss: -1952.163281\n",
            "Train Epoch: 35 [28160/60000 (47%)]\tLoss: -1985.210547\n",
            "Train Epoch: 35 [29440/60000 (49%)]\tLoss: -1939.972852\n",
            "Train Epoch: 35 [30720/60000 (51%)]\tLoss: -1949.704883\n",
            "Train Epoch: 35 [32000/60000 (53%)]\tLoss: -1940.560547\n",
            "Train Epoch: 35 [33280/60000 (55%)]\tLoss: -1928.827930\n",
            "Train Epoch: 35 [34560/60000 (58%)]\tLoss: -1917.438477\n",
            "Train Epoch: 35 [35840/60000 (60%)]\tLoss: -2002.815039\n",
            "Train Epoch: 35 [37120/60000 (62%)]\tLoss: -2053.531055\n",
            "Train Epoch: 35 [38400/60000 (64%)]\tLoss: -1964.518359\n",
            "Train Epoch: 35 [39680/60000 (66%)]\tLoss: -1965.142578\n",
            "Train Epoch: 35 [40960/60000 (68%)]\tLoss: -2007.507227\n",
            "Train Epoch: 35 [42240/60000 (70%)]\tLoss: -1950.959570\n",
            "Train Epoch: 35 [43520/60000 (72%)]\tLoss: -1944.453906\n",
            "Train Epoch: 35 [44800/60000 (75%)]\tLoss: -2010.805078\n",
            "Train Epoch: 35 [46080/60000 (77%)]\tLoss: -1979.272266\n",
            "Train Epoch: 35 [47360/60000 (79%)]\tLoss: -1965.537891\n",
            "Train Epoch: 35 [48640/60000 (81%)]\tLoss: -1980.970117\n",
            "Train Epoch: 35 [49920/60000 (83%)]\tLoss: -2000.729883\n",
            "Train Epoch: 35 [51200/60000 (85%)]\tLoss: -2026.445703\n",
            "Train Epoch: 35 [52480/60000 (87%)]\tLoss: -1975.856641\n",
            "Train Epoch: 35 [53760/60000 (90%)]\tLoss: -1957.960352\n",
            "Train Epoch: 35 [55040/60000 (92%)]\tLoss: -1933.297461\n",
            "Train Epoch: 35 [56320/60000 (94%)]\tLoss: -1953.707813\n",
            "Train Epoch: 35 [57600/60000 (96%)]\tLoss: -1963.101172\n",
            "Train Epoch: 35 [58880/60000 (98%)]\tLoss: -1917.858984\n",
            "====> Epoch: 35 Average loss: -19682.0142\n",
            "Train Epoch: 36 [0/60000 (0%)]\tLoss: -19352.939453\n",
            "Train Epoch: 36 [1280/60000 (2%)]\tLoss: -2011.106641\n",
            "Train Epoch: 36 [2560/60000 (4%)]\tLoss: -1926.220703\n",
            "Train Epoch: 36 [3840/60000 (6%)]\tLoss: -1954.632031\n",
            "Train Epoch: 36 [5120/60000 (9%)]\tLoss: -1964.052734\n",
            "Train Epoch: 36 [6400/60000 (11%)]\tLoss: -1932.737695\n",
            "Train Epoch: 36 [7680/60000 (13%)]\tLoss: -1952.369727\n",
            "Train Epoch: 36 [8960/60000 (15%)]\tLoss: -1996.953516\n",
            "Train Epoch: 36 [10240/60000 (17%)]\tLoss: -2161.147852\n",
            "Train Epoch: 36 [11520/60000 (19%)]\tLoss: -1962.224414\n",
            "Train Epoch: 36 [12800/60000 (21%)]\tLoss: -1954.932031\n",
            "Train Epoch: 36 [14080/60000 (23%)]\tLoss: -2065.406250\n",
            "Train Epoch: 36 [15360/60000 (26%)]\tLoss: -1957.719727\n",
            "Train Epoch: 36 [16640/60000 (28%)]\tLoss: -1953.281836\n",
            "Train Epoch: 36 [17920/60000 (30%)]\tLoss: -1950.875781\n",
            "Train Epoch: 36 [19200/60000 (32%)]\tLoss: -1997.000781\n",
            "Train Epoch: 36 [20480/60000 (34%)]\tLoss: -1906.747461\n",
            "Train Epoch: 36 [21760/60000 (36%)]\tLoss: -2022.174023\n",
            "Train Epoch: 36 [23040/60000 (38%)]\tLoss: -1992.309180\n",
            "Train Epoch: 36 [24320/60000 (41%)]\tLoss: -2040.487500\n",
            "Train Epoch: 36 [25600/60000 (43%)]\tLoss: -1916.928320\n",
            "Train Epoch: 36 [26880/60000 (45%)]\tLoss: -1950.643945\n",
            "Train Epoch: 36 [28160/60000 (47%)]\tLoss: -2003.489844\n",
            "Train Epoch: 36 [29440/60000 (49%)]\tLoss: -1946.013672\n",
            "Train Epoch: 36 [30720/60000 (51%)]\tLoss: -1969.821875\n",
            "Train Epoch: 36 [32000/60000 (53%)]\tLoss: -1956.551563\n",
            "Train Epoch: 36 [33280/60000 (55%)]\tLoss: -1926.946289\n",
            "Train Epoch: 36 [34560/60000 (58%)]\tLoss: -1925.772266\n",
            "Train Epoch: 36 [35840/60000 (60%)]\tLoss: -1981.286133\n",
            "Train Epoch: 36 [37120/60000 (62%)]\tLoss: -2029.576367\n",
            "Train Epoch: 36 [38400/60000 (64%)]\tLoss: -1963.482031\n",
            "Train Epoch: 36 [39680/60000 (66%)]\tLoss: -1959.041211\n",
            "Train Epoch: 36 [40960/60000 (68%)]\tLoss: -2022.426758\n",
            "Train Epoch: 36 [42240/60000 (70%)]\tLoss: -1963.415039\n",
            "Train Epoch: 36 [43520/60000 (72%)]\tLoss: -1946.202930\n",
            "Train Epoch: 36 [44800/60000 (75%)]\tLoss: -2024.318359\n",
            "Train Epoch: 36 [46080/60000 (77%)]\tLoss: -1964.616211\n",
            "Train Epoch: 36 [47360/60000 (79%)]\tLoss: -1972.737891\n",
            "Train Epoch: 36 [48640/60000 (81%)]\tLoss: -1962.119531\n",
            "Train Epoch: 36 [49920/60000 (83%)]\tLoss: -1992.873047\n",
            "Train Epoch: 36 [51200/60000 (85%)]\tLoss: -1964.969727\n",
            "Train Epoch: 36 [52480/60000 (87%)]\tLoss: -1976.808398\n",
            "Train Epoch: 36 [53760/60000 (90%)]\tLoss: -1960.432617\n",
            "Train Epoch: 36 [55040/60000 (92%)]\tLoss: -1932.940625\n",
            "Train Epoch: 36 [56320/60000 (94%)]\tLoss: -1951.713672\n",
            "Train Epoch: 36 [57600/60000 (96%)]\tLoss: -1958.666406\n",
            "Train Epoch: 36 [58880/60000 (98%)]\tLoss: -1917.331250\n",
            "====> Epoch: 36 Average loss: -19702.9020\n",
            "Train Epoch: 37 [0/60000 (0%)]\tLoss: -19427.615234\n",
            "Train Epoch: 37 [1280/60000 (2%)]\tLoss: -2033.724414\n",
            "Train Epoch: 37 [2560/60000 (4%)]\tLoss: -1925.389453\n",
            "Train Epoch: 37 [3840/60000 (6%)]\tLoss: -1962.120312\n",
            "Train Epoch: 37 [5120/60000 (9%)]\tLoss: -1971.789453\n",
            "Train Epoch: 37 [6400/60000 (11%)]\tLoss: -1931.511328\n",
            "Train Epoch: 37 [7680/60000 (13%)]\tLoss: -1948.978125\n",
            "Train Epoch: 37 [8960/60000 (15%)]\tLoss: -1991.416211\n",
            "Train Epoch: 37 [10240/60000 (17%)]\tLoss: -2209.537500\n",
            "Train Epoch: 37 [11520/60000 (19%)]\tLoss: -1961.947852\n",
            "Train Epoch: 37 [12800/60000 (21%)]\tLoss: -1946.759570\n",
            "Train Epoch: 37 [14080/60000 (23%)]\tLoss: -2109.744336\n",
            "Train Epoch: 37 [15360/60000 (26%)]\tLoss: -1957.383594\n",
            "Train Epoch: 37 [16640/60000 (28%)]\tLoss: -1952.256055\n",
            "Train Epoch: 37 [17920/60000 (30%)]\tLoss: -1947.512305\n",
            "Train Epoch: 37 [19200/60000 (32%)]\tLoss: -2018.280664\n",
            "Train Epoch: 37 [20480/60000 (34%)]\tLoss: -1912.543945\n",
            "Train Epoch: 37 [21760/60000 (36%)]\tLoss: -1981.014062\n",
            "Train Epoch: 37 [23040/60000 (38%)]\tLoss: -1986.230469\n",
            "Train Epoch: 37 [24320/60000 (41%)]\tLoss: -2013.072070\n",
            "Train Epoch: 37 [25600/60000 (43%)]\tLoss: -1926.098047\n",
            "Train Epoch: 37 [26880/60000 (45%)]\tLoss: -1950.320312\n",
            "Train Epoch: 37 [28160/60000 (47%)]\tLoss: -2013.388867\n",
            "Train Epoch: 37 [29440/60000 (49%)]\tLoss: -1936.373437\n",
            "Train Epoch: 37 [30720/60000 (51%)]\tLoss: -1960.206641\n",
            "Train Epoch: 37 [32000/60000 (53%)]\tLoss: -1943.789258\n",
            "Train Epoch: 37 [33280/60000 (55%)]\tLoss: -1929.862695\n",
            "Train Epoch: 37 [34560/60000 (58%)]\tLoss: -1927.032227\n",
            "Train Epoch: 37 [35840/60000 (60%)]\tLoss: -1997.371289\n",
            "Train Epoch: 37 [37120/60000 (62%)]\tLoss: -2027.897266\n",
            "Train Epoch: 37 [38400/60000 (64%)]\tLoss: -1958.460547\n",
            "Train Epoch: 37 [39680/60000 (66%)]\tLoss: -1967.260937\n",
            "Train Epoch: 37 [40960/60000 (68%)]\tLoss: -2026.920312\n",
            "Train Epoch: 37 [42240/60000 (70%)]\tLoss: -1974.916016\n",
            "Train Epoch: 37 [43520/60000 (72%)]\tLoss: -1944.981836\n",
            "Train Epoch: 37 [44800/60000 (75%)]\tLoss: -2011.490430\n",
            "Train Epoch: 37 [46080/60000 (77%)]\tLoss: -1989.903711\n",
            "Train Epoch: 37 [47360/60000 (79%)]\tLoss: -1972.884180\n",
            "Train Epoch: 37 [48640/60000 (81%)]\tLoss: -1960.907422\n",
            "Train Epoch: 37 [49920/60000 (83%)]\tLoss: -1994.303516\n",
            "Train Epoch: 37 [51200/60000 (85%)]\tLoss: -1984.670898\n",
            "Train Epoch: 37 [52480/60000 (87%)]\tLoss: -1972.799219\n",
            "Train Epoch: 37 [53760/60000 (90%)]\tLoss: -1955.599805\n",
            "Train Epoch: 37 [55040/60000 (92%)]\tLoss: -1911.295703\n",
            "Train Epoch: 37 [56320/60000 (94%)]\tLoss: -1963.317969\n",
            "Train Epoch: 37 [57600/60000 (96%)]\tLoss: -1960.640234\n",
            "Train Epoch: 37 [58880/60000 (98%)]\tLoss: -1919.451367\n",
            "====> Epoch: 37 Average loss: -19716.6549\n",
            "Train Epoch: 38 [0/60000 (0%)]\tLoss: -19326.291016\n",
            "Train Epoch: 38 [1280/60000 (2%)]\tLoss: -1981.682031\n",
            "Train Epoch: 38 [2560/60000 (4%)]\tLoss: -1930.377930\n",
            "Train Epoch: 38 [3840/60000 (6%)]\tLoss: -1933.752734\n",
            "Train Epoch: 38 [5120/60000 (9%)]\tLoss: -1975.854687\n",
            "Train Epoch: 38 [6400/60000 (11%)]\tLoss: -1929.184180\n",
            "Train Epoch: 38 [7680/60000 (13%)]\tLoss: -1954.620898\n",
            "Train Epoch: 38 [8960/60000 (15%)]\tLoss: -1976.651758\n",
            "Train Epoch: 38 [10240/60000 (17%)]\tLoss: -2189.037695\n",
            "Train Epoch: 38 [11520/60000 (19%)]\tLoss: -1956.341602\n",
            "Train Epoch: 38 [12800/60000 (21%)]\tLoss: -1959.397461\n",
            "Train Epoch: 38 [14080/60000 (23%)]\tLoss: -2199.206836\n",
            "Train Epoch: 38 [15360/60000 (26%)]\tLoss: -1952.707813\n",
            "Train Epoch: 38 [16640/60000 (28%)]\tLoss: -1953.177148\n",
            "Train Epoch: 38 [17920/60000 (30%)]\tLoss: -1949.990430\n",
            "Train Epoch: 38 [19200/60000 (32%)]\tLoss: -2012.807617\n",
            "Train Epoch: 38 [20480/60000 (34%)]\tLoss: -1916.446875\n",
            "Train Epoch: 38 [21760/60000 (36%)]\tLoss: -1988.276367\n",
            "Train Epoch: 38 [23040/60000 (38%)]\tLoss: -2007.932227\n",
            "Train Epoch: 38 [24320/60000 (41%)]\tLoss: -2033.732031\n",
            "Train Epoch: 38 [25600/60000 (43%)]\tLoss: -1916.356055\n",
            "Train Epoch: 38 [26880/60000 (45%)]\tLoss: -1948.663867\n",
            "Train Epoch: 38 [28160/60000 (47%)]\tLoss: -2021.455859\n",
            "Train Epoch: 38 [29440/60000 (49%)]\tLoss: -1935.069531\n",
            "Train Epoch: 38 [30720/60000 (51%)]\tLoss: -1944.475000\n",
            "Train Epoch: 38 [32000/60000 (53%)]\tLoss: -1942.183008\n",
            "Train Epoch: 38 [33280/60000 (55%)]\tLoss: -1925.399023\n",
            "Train Epoch: 38 [34560/60000 (58%)]\tLoss: -1925.892188\n",
            "Train Epoch: 38 [35840/60000 (60%)]\tLoss: -1991.898828\n",
            "Train Epoch: 38 [37120/60000 (62%)]\tLoss: -2034.495703\n",
            "Train Epoch: 38 [38400/60000 (64%)]\tLoss: -1962.369922\n",
            "Train Epoch: 38 [39680/60000 (66%)]\tLoss: -1961.132031\n",
            "Train Epoch: 38 [40960/60000 (68%)]\tLoss: -2030.087891\n",
            "Train Epoch: 38 [42240/60000 (70%)]\tLoss: -1944.696484\n",
            "Train Epoch: 38 [43520/60000 (72%)]\tLoss: -1943.646289\n",
            "Train Epoch: 38 [44800/60000 (75%)]\tLoss: -2016.714453\n",
            "Train Epoch: 38 [46080/60000 (77%)]\tLoss: -1975.858594\n",
            "Train Epoch: 38 [47360/60000 (79%)]\tLoss: -1963.836328\n",
            "Train Epoch: 38 [48640/60000 (81%)]\tLoss: -1964.448633\n",
            "Train Epoch: 38 [49920/60000 (83%)]\tLoss: -1988.624219\n",
            "Train Epoch: 38 [51200/60000 (85%)]\tLoss: -1994.812305\n",
            "Train Epoch: 38 [52480/60000 (87%)]\tLoss: -1975.566406\n",
            "Train Epoch: 38 [53760/60000 (90%)]\tLoss: -1955.890039\n",
            "Train Epoch: 38 [55040/60000 (92%)]\tLoss: -1933.261719\n",
            "Train Epoch: 38 [56320/60000 (94%)]\tLoss: -1953.812109\n",
            "Train Epoch: 38 [57600/60000 (96%)]\tLoss: -1962.801563\n",
            "Train Epoch: 38 [58880/60000 (98%)]\tLoss: -1916.816211\n",
            "====> Epoch: 38 Average loss: -19723.1354\n",
            "Train Epoch: 39 [0/60000 (0%)]\tLoss: -19380.476562\n",
            "Train Epoch: 39 [1280/60000 (2%)]\tLoss: -2033.257422\n",
            "Train Epoch: 39 [2560/60000 (4%)]\tLoss: -1927.452539\n",
            "Train Epoch: 39 [3840/60000 (6%)]\tLoss: -1945.562891\n",
            "Train Epoch: 39 [5120/60000 (9%)]\tLoss: -1968.941406\n",
            "Train Epoch: 39 [6400/60000 (11%)]\tLoss: -1930.258203\n",
            "Train Epoch: 39 [7680/60000 (13%)]\tLoss: -1946.941211\n",
            "Train Epoch: 39 [8960/60000 (15%)]\tLoss: -1976.911328\n",
            "Train Epoch: 39 [10240/60000 (17%)]\tLoss: -2220.291992\n",
            "Train Epoch: 39 [11520/60000 (19%)]\tLoss: -1956.505859\n",
            "Train Epoch: 39 [12800/60000 (21%)]\tLoss: -1973.500391\n",
            "Train Epoch: 39 [14080/60000 (23%)]\tLoss: -2145.049219\n",
            "Train Epoch: 39 [15360/60000 (26%)]\tLoss: -1953.301758\n",
            "Train Epoch: 39 [16640/60000 (28%)]\tLoss: -1951.768945\n",
            "Train Epoch: 39 [17920/60000 (30%)]\tLoss: -1947.667383\n",
            "Train Epoch: 39 [19200/60000 (32%)]\tLoss: -2019.282031\n",
            "Train Epoch: 39 [20480/60000 (34%)]\tLoss: -1923.074414\n",
            "Train Epoch: 39 [21760/60000 (36%)]\tLoss: -2006.998242\n",
            "Train Epoch: 39 [23040/60000 (38%)]\tLoss: -2010.642773\n",
            "Train Epoch: 39 [24320/60000 (41%)]\tLoss: -2005.637109\n",
            "Train Epoch: 39 [25600/60000 (43%)]\tLoss: -1917.795508\n",
            "Train Epoch: 39 [26880/60000 (45%)]\tLoss: -1950.055859\n",
            "Train Epoch: 39 [28160/60000 (47%)]\tLoss: -2010.097852\n",
            "Train Epoch: 39 [29440/60000 (49%)]\tLoss: -1944.155078\n",
            "Train Epoch: 39 [30720/60000 (51%)]\tLoss: -1954.570703\n",
            "Train Epoch: 39 [32000/60000 (53%)]\tLoss: -1945.934961\n",
            "Train Epoch: 39 [33280/60000 (55%)]\tLoss: -1926.415430\n",
            "Train Epoch: 39 [34560/60000 (58%)]\tLoss: -1929.354492\n",
            "Train Epoch: 39 [35840/60000 (60%)]\tLoss: -1980.469336\n",
            "Train Epoch: 39 [37120/60000 (62%)]\tLoss: -2031.184961\n",
            "Train Epoch: 39 [38400/60000 (64%)]\tLoss: -1956.031250\n",
            "Train Epoch: 39 [39680/60000 (66%)]\tLoss: -1968.326562\n",
            "Train Epoch: 39 [40960/60000 (68%)]\tLoss: -2005.983398\n",
            "Train Epoch: 39 [42240/60000 (70%)]\tLoss: -1953.807812\n",
            "Train Epoch: 39 [43520/60000 (72%)]\tLoss: -1943.767773\n",
            "Train Epoch: 39 [44800/60000 (75%)]\tLoss: -2016.278125\n",
            "Train Epoch: 39 [46080/60000 (77%)]\tLoss: -1980.176172\n",
            "Train Epoch: 39 [47360/60000 (79%)]\tLoss: -1968.039648\n",
            "Train Epoch: 39 [48640/60000 (81%)]\tLoss: -1961.858789\n",
            "Train Epoch: 39 [49920/60000 (83%)]\tLoss: -1997.653516\n",
            "Train Epoch: 39 [51200/60000 (85%)]\tLoss: -2024.731445\n",
            "Train Epoch: 39 [52480/60000 (87%)]\tLoss: -1971.778711\n",
            "Train Epoch: 39 [53760/60000 (90%)]\tLoss: -1955.509180\n",
            "Train Epoch: 39 [55040/60000 (92%)]\tLoss: -1927.710156\n",
            "Train Epoch: 39 [56320/60000 (94%)]\tLoss: -1955.140430\n",
            "Train Epoch: 39 [57600/60000 (96%)]\tLoss: -1959.864258\n",
            "Train Epoch: 39 [58880/60000 (98%)]\tLoss: -1914.708203\n",
            "====> Epoch: 39 Average loss: -19730.3011\n",
            "Train Epoch: 40 [0/60000 (0%)]\tLoss: -19208.367188\n",
            "Train Epoch: 40 [1280/60000 (2%)]\tLoss: -2037.941992\n",
            "Train Epoch: 40 [2560/60000 (4%)]\tLoss: -1929.159961\n",
            "Train Epoch: 40 [3840/60000 (6%)]\tLoss: -1962.267773\n",
            "Train Epoch: 40 [5120/60000 (9%)]\tLoss: -1979.877539\n",
            "Train Epoch: 40 [6400/60000 (11%)]\tLoss: -1931.376367\n",
            "Train Epoch: 40 [7680/60000 (13%)]\tLoss: -1945.218945\n",
            "Train Epoch: 40 [8960/60000 (15%)]\tLoss: -1998.797852\n",
            "Train Epoch: 40 [10240/60000 (17%)]\tLoss: -2210.352930\n",
            "Train Epoch: 40 [11520/60000 (19%)]\tLoss: -1957.957031\n",
            "Train Epoch: 40 [12800/60000 (21%)]\tLoss: -1981.470117\n",
            "Train Epoch: 40 [14080/60000 (23%)]\tLoss: -2127.578320\n",
            "Train Epoch: 40 [15360/60000 (26%)]\tLoss: -1950.030078\n",
            "Train Epoch: 40 [16640/60000 (28%)]\tLoss: -1955.504883\n",
            "Train Epoch: 40 [17920/60000 (30%)]\tLoss: -1947.304492\n",
            "Train Epoch: 40 [19200/60000 (32%)]\tLoss: -2008.751758\n",
            "Train Epoch: 40 [20480/60000 (34%)]\tLoss: -1923.450586\n",
            "Train Epoch: 40 [21760/60000 (36%)]\tLoss: -2047.388086\n",
            "Train Epoch: 40 [23040/60000 (38%)]\tLoss: -2006.773242\n",
            "Train Epoch: 40 [24320/60000 (41%)]\tLoss: -2023.262305\n",
            "Train Epoch: 40 [25600/60000 (43%)]\tLoss: -1954.134570\n",
            "Train Epoch: 40 [26880/60000 (45%)]\tLoss: -1952.196289\n",
            "Train Epoch: 40 [28160/60000 (47%)]\tLoss: -2007.235156\n",
            "Train Epoch: 40 [29440/60000 (49%)]\tLoss: -1955.256055\n",
            "Train Epoch: 40 [30720/60000 (51%)]\tLoss: -1966.813086\n",
            "Train Epoch: 40 [32000/60000 (53%)]\tLoss: -1955.808984\n",
            "Train Epoch: 40 [33280/60000 (55%)]\tLoss: -1920.645313\n",
            "Train Epoch: 40 [34560/60000 (58%)]\tLoss: -1928.612891\n",
            "Train Epoch: 40 [35840/60000 (60%)]\tLoss: -1976.785547\n",
            "Train Epoch: 40 [37120/60000 (62%)]\tLoss: -2039.424805\n",
            "Train Epoch: 40 [38400/60000 (64%)]\tLoss: -1978.125781\n",
            "Train Epoch: 40 [39680/60000 (66%)]\tLoss: -1960.373047\n",
            "Train Epoch: 40 [40960/60000 (68%)]\tLoss: -2031.692969\n",
            "Train Epoch: 40 [42240/60000 (70%)]\tLoss: -1964.943750\n",
            "Train Epoch: 40 [43520/60000 (72%)]\tLoss: -1940.315234\n",
            "Train Epoch: 40 [44800/60000 (75%)]\tLoss: -2024.650391\n",
            "Train Epoch: 40 [46080/60000 (77%)]\tLoss: -1979.969727\n",
            "Train Epoch: 40 [47360/60000 (79%)]\tLoss: -1977.267578\n",
            "Train Epoch: 40 [48640/60000 (81%)]\tLoss: -1960.191211\n",
            "Train Epoch: 40 [49920/60000 (83%)]\tLoss: -2021.052734\n",
            "Train Epoch: 40 [51200/60000 (85%)]\tLoss: -1973.308594\n",
            "Train Epoch: 40 [52480/60000 (87%)]\tLoss: -1979.384375\n",
            "Train Epoch: 40 [53760/60000 (90%)]\tLoss: -1964.886914\n",
            "Train Epoch: 40 [55040/60000 (92%)]\tLoss: -1908.876953\n",
            "Train Epoch: 40 [56320/60000 (94%)]\tLoss: -1967.348828\n",
            "Train Epoch: 40 [57600/60000 (96%)]\tLoss: -1955.121094\n",
            "Train Epoch: 40 [58880/60000 (98%)]\tLoss: -1911.905469\n",
            "====> Epoch: 40 Average loss: -19778.9016\n",
            "Train Epoch: 41 [0/60000 (0%)]\tLoss: -19280.652344\n",
            "Train Epoch: 41 [1280/60000 (2%)]\tLoss: -2012.330859\n",
            "Train Epoch: 41 [2560/60000 (4%)]\tLoss: -1958.630664\n",
            "Train Epoch: 41 [3840/60000 (6%)]\tLoss: -1952.050977\n",
            "Train Epoch: 41 [5120/60000 (9%)]\tLoss: -2001.383984\n",
            "Train Epoch: 41 [6400/60000 (11%)]\tLoss: -1931.188281\n",
            "Train Epoch: 41 [7680/60000 (13%)]\tLoss: -1965.632422\n",
            "Train Epoch: 41 [8960/60000 (15%)]\tLoss: -1986.685156\n",
            "Train Epoch: 41 [10240/60000 (17%)]\tLoss: -2207.373828\n",
            "Train Epoch: 41 [11520/60000 (19%)]\tLoss: -1951.309766\n",
            "Train Epoch: 41 [12800/60000 (21%)]\tLoss: -1959.880078\n",
            "Train Epoch: 41 [14080/60000 (23%)]\tLoss: -2175.224023\n",
            "Train Epoch: 41 [15360/60000 (26%)]\tLoss: -1949.984375\n",
            "Train Epoch: 41 [16640/60000 (28%)]\tLoss: -1952.992383\n",
            "Train Epoch: 41 [17920/60000 (30%)]\tLoss: -1948.844727\n",
            "Train Epoch: 41 [19200/60000 (32%)]\tLoss: -2008.786914\n",
            "Train Epoch: 41 [20480/60000 (34%)]\tLoss: -1921.898828\n",
            "Train Epoch: 41 [21760/60000 (36%)]\tLoss: -2041.126172\n",
            "Train Epoch: 41 [23040/60000 (38%)]\tLoss: -2061.245508\n",
            "Train Epoch: 41 [24320/60000 (41%)]\tLoss: -2027.341211\n",
            "Train Epoch: 41 [25600/60000 (43%)]\tLoss: -1945.521680\n",
            "Train Epoch: 41 [26880/60000 (45%)]\tLoss: -1964.583984\n",
            "Train Epoch: 41 [28160/60000 (47%)]\tLoss: -2035.981445\n",
            "Train Epoch: 41 [29440/60000 (49%)]\tLoss: -1951.823828\n",
            "Train Epoch: 41 [30720/60000 (51%)]\tLoss: -1953.623437\n",
            "Train Epoch: 41 [32000/60000 (53%)]\tLoss: -1963.107227\n",
            "Train Epoch: 41 [33280/60000 (55%)]\tLoss: -1918.302344\n",
            "Train Epoch: 41 [34560/60000 (58%)]\tLoss: -1937.121484\n",
            "Train Epoch: 41 [35840/60000 (60%)]\tLoss: -1996.769922\n",
            "Train Epoch: 41 [37120/60000 (62%)]\tLoss: -2103.040625\n",
            "Train Epoch: 41 [38400/60000 (64%)]\tLoss: -1985.158789\n",
            "Train Epoch: 41 [39680/60000 (66%)]\tLoss: -1967.537109\n",
            "Train Epoch: 41 [40960/60000 (68%)]\tLoss: -2047.156641\n",
            "Train Epoch: 41 [42240/60000 (70%)]\tLoss: -1963.018750\n",
            "Train Epoch: 41 [43520/60000 (72%)]\tLoss: -1953.132617\n",
            "Train Epoch: 41 [44800/60000 (75%)]\tLoss: -2007.658984\n",
            "Train Epoch: 41 [46080/60000 (77%)]\tLoss: -2017.663477\n",
            "Train Epoch: 41 [47360/60000 (79%)]\tLoss: -1976.491211\n",
            "Train Epoch: 41 [48640/60000 (81%)]\tLoss: -1970.129297\n",
            "Train Epoch: 41 [49920/60000 (83%)]\tLoss: -1997.090430\n",
            "Train Epoch: 41 [51200/60000 (85%)]\tLoss: -1974.886328\n",
            "Train Epoch: 41 [52480/60000 (87%)]\tLoss: -1971.673047\n",
            "Train Epoch: 41 [53760/60000 (90%)]\tLoss: -1958.174805\n",
            "Train Epoch: 41 [55040/60000 (92%)]\tLoss: -1931.510547\n",
            "Train Epoch: 41 [56320/60000 (94%)]\tLoss: -1966.034180\n",
            "Train Epoch: 41 [57600/60000 (96%)]\tLoss: -1964.426172\n",
            "Train Epoch: 41 [58880/60000 (98%)]\tLoss: -1913.242383\n",
            "====> Epoch: 41 Average loss: -19792.7521\n",
            "Train Epoch: 42 [0/60000 (0%)]\tLoss: -19356.142578\n",
            "Train Epoch: 42 [1280/60000 (2%)]\tLoss: -2006.782617\n",
            "Train Epoch: 42 [2560/60000 (4%)]\tLoss: -1929.675391\n",
            "Train Epoch: 42 [3840/60000 (6%)]\tLoss: -1960.087305\n",
            "Train Epoch: 42 [5120/60000 (9%)]\tLoss: -1989.861133\n",
            "Train Epoch: 42 [6400/60000 (11%)]\tLoss: -1933.501367\n",
            "Train Epoch: 42 [7680/60000 (13%)]\tLoss: -1947.032617\n",
            "Train Epoch: 42 [8960/60000 (15%)]\tLoss: -1977.235156\n",
            "Train Epoch: 42 [10240/60000 (17%)]\tLoss: -2230.957812\n",
            "Train Epoch: 42 [11520/60000 (19%)]\tLoss: -1951.979102\n",
            "Train Epoch: 42 [12800/60000 (21%)]\tLoss: -1946.225586\n",
            "Train Epoch: 42 [14080/60000 (23%)]\tLoss: -2156.329102\n",
            "Train Epoch: 42 [15360/60000 (26%)]\tLoss: -1948.821484\n",
            "Train Epoch: 42 [16640/60000 (28%)]\tLoss: -1953.001953\n",
            "Train Epoch: 42 [17920/60000 (30%)]\tLoss: -1956.183008\n",
            "Train Epoch: 42 [19200/60000 (32%)]\tLoss: -1989.567773\n",
            "Train Epoch: 42 [20480/60000 (34%)]\tLoss: -1921.152734\n",
            "Train Epoch: 42 [21760/60000 (36%)]\tLoss: -1991.581641\n",
            "Train Epoch: 42 [23040/60000 (38%)]\tLoss: -2047.063281\n",
            "Train Epoch: 42 [24320/60000 (41%)]\tLoss: -2040.146094\n",
            "Train Epoch: 42 [25600/60000 (43%)]\tLoss: -1927.076758\n",
            "Train Epoch: 42 [26880/60000 (45%)]\tLoss: -1952.228711\n",
            "Train Epoch: 42 [28160/60000 (47%)]\tLoss: -1993.525000\n",
            "Train Epoch: 42 [29440/60000 (49%)]\tLoss: -1936.611523\n",
            "Train Epoch: 42 [30720/60000 (51%)]\tLoss: -1952.569922\n",
            "Train Epoch: 42 [32000/60000 (53%)]\tLoss: -1949.989258\n",
            "Train Epoch: 42 [33280/60000 (55%)]\tLoss: -1922.794336\n",
            "Train Epoch: 42 [34560/60000 (58%)]\tLoss: -1928.722266\n",
            "Train Epoch: 42 [35840/60000 (60%)]\tLoss: -2019.308203\n",
            "Train Epoch: 42 [37120/60000 (62%)]\tLoss: -2073.780664\n",
            "Train Epoch: 42 [38400/60000 (64%)]\tLoss: -1972.706836\n",
            "Train Epoch: 42 [39680/60000 (66%)]\tLoss: -1958.399414\n",
            "Train Epoch: 42 [40960/60000 (68%)]\tLoss: -2011.759766\n",
            "Train Epoch: 42 [42240/60000 (70%)]\tLoss: -1941.954297\n",
            "Train Epoch: 42 [43520/60000 (72%)]\tLoss: -1936.797852\n",
            "Train Epoch: 42 [44800/60000 (75%)]\tLoss: -2000.426367\n",
            "Train Epoch: 42 [46080/60000 (77%)]\tLoss: -1975.417187\n",
            "Train Epoch: 42 [47360/60000 (79%)]\tLoss: -1998.534766\n",
            "Train Epoch: 42 [48640/60000 (81%)]\tLoss: -1958.635742\n",
            "Train Epoch: 42 [49920/60000 (83%)]\tLoss: -2024.170508\n",
            "Train Epoch: 42 [51200/60000 (85%)]\tLoss: -1986.360156\n",
            "Train Epoch: 42 [52480/60000 (87%)]\tLoss: -1963.787305\n",
            "Train Epoch: 42 [53760/60000 (90%)]\tLoss: -1957.643164\n",
            "Train Epoch: 42 [55040/60000 (92%)]\tLoss: -1911.978516\n",
            "Train Epoch: 42 [56320/60000 (94%)]\tLoss: -1945.785352\n",
            "Train Epoch: 42 [57600/60000 (96%)]\tLoss: -1963.742578\n",
            "Train Epoch: 42 [58880/60000 (98%)]\tLoss: -1914.603125\n",
            "====> Epoch: 42 Average loss: -19748.3853\n",
            "Train Epoch: 43 [0/60000 (0%)]\tLoss: -19443.390625\n",
            "Train Epoch: 43 [1280/60000 (2%)]\tLoss: -2099.795117\n",
            "Train Epoch: 43 [2560/60000 (4%)]\tLoss: -1944.157617\n",
            "Train Epoch: 43 [3840/60000 (6%)]\tLoss: -1970.354687\n",
            "Train Epoch: 43 [5120/60000 (9%)]\tLoss: -1967.539648\n",
            "Train Epoch: 43 [6400/60000 (11%)]\tLoss: -1932.668945\n",
            "Train Epoch: 43 [7680/60000 (13%)]\tLoss: -1975.644336\n",
            "Train Epoch: 43 [8960/60000 (15%)]\tLoss: -2010.958789\n",
            "Train Epoch: 43 [10240/60000 (17%)]\tLoss: -2252.255273\n",
            "Train Epoch: 43 [11520/60000 (19%)]\tLoss: -1952.022852\n",
            "Train Epoch: 43 [12800/60000 (21%)]\tLoss: -1972.147070\n",
            "Train Epoch: 43 [14080/60000 (23%)]\tLoss: -2200.414648\n",
            "Train Epoch: 43 [15360/60000 (26%)]\tLoss: -1949.666797\n",
            "Train Epoch: 43 [16640/60000 (28%)]\tLoss: -1949.150586\n",
            "Train Epoch: 43 [17920/60000 (30%)]\tLoss: -1951.137500\n",
            "Train Epoch: 43 [19200/60000 (32%)]\tLoss: -1989.731641\n",
            "Train Epoch: 43 [20480/60000 (34%)]\tLoss: -1933.175586\n",
            "Train Epoch: 43 [21760/60000 (36%)]\tLoss: -2020.629883\n",
            "Train Epoch: 43 [23040/60000 (38%)]\tLoss: -2060.850586\n",
            "Train Epoch: 43 [24320/60000 (41%)]\tLoss: -2046.243555\n",
            "Train Epoch: 43 [25600/60000 (43%)]\tLoss: -1908.119727\n",
            "Train Epoch: 43 [26880/60000 (45%)]\tLoss: -1957.172656\n",
            "Train Epoch: 43 [28160/60000 (47%)]\tLoss: -2010.669336\n",
            "Train Epoch: 43 [29440/60000 (49%)]\tLoss: -1932.568164\n",
            "Train Epoch: 43 [30720/60000 (51%)]\tLoss: -1956.676367\n",
            "Train Epoch: 43 [32000/60000 (53%)]\tLoss: -1982.312695\n",
            "Train Epoch: 43 [33280/60000 (55%)]\tLoss: -1922.560156\n",
            "Train Epoch: 43 [34560/60000 (58%)]\tLoss: -1921.421289\n",
            "Train Epoch: 43 [35840/60000 (60%)]\tLoss: -2013.575195\n",
            "Train Epoch: 43 [37120/60000 (62%)]\tLoss: -2100.644141\n",
            "Train Epoch: 43 [38400/60000 (64%)]\tLoss: -1974.514062\n",
            "Train Epoch: 43 [39680/60000 (66%)]\tLoss: -1962.560352\n",
            "Train Epoch: 43 [40960/60000 (68%)]\tLoss: -2035.701758\n",
            "Train Epoch: 43 [42240/60000 (70%)]\tLoss: -1949.806641\n",
            "Train Epoch: 43 [43520/60000 (72%)]\tLoss: -1937.238281\n",
            "Train Epoch: 43 [44800/60000 (75%)]\tLoss: -2004.173633\n",
            "Train Epoch: 43 [46080/60000 (77%)]\tLoss: -1996.156445\n",
            "Train Epoch: 43 [47360/60000 (79%)]\tLoss: -2004.224414\n",
            "Train Epoch: 43 [48640/60000 (81%)]\tLoss: -1994.604492\n",
            "Train Epoch: 43 [49920/60000 (83%)]\tLoss: -1995.515625\n",
            "Train Epoch: 43 [51200/60000 (85%)]\tLoss: -2034.128711\n",
            "Train Epoch: 43 [52480/60000 (87%)]\tLoss: -1966.133203\n",
            "Train Epoch: 43 [53760/60000 (90%)]\tLoss: -1957.006836\n",
            "Train Epoch: 43 [55040/60000 (92%)]\tLoss: -1927.747070\n",
            "Train Epoch: 43 [56320/60000 (94%)]\tLoss: -1960.748437\n",
            "Train Epoch: 43 [57600/60000 (96%)]\tLoss: -1959.004883\n",
            "Train Epoch: 43 [58880/60000 (98%)]\tLoss: -1913.151758\n",
            "====> Epoch: 43 Average loss: -19774.5829\n",
            "Train Epoch: 44 [0/60000 (0%)]\tLoss: -19373.240234\n",
            "Train Epoch: 44 [1280/60000 (2%)]\tLoss: -2080.159766\n",
            "Train Epoch: 44 [2560/60000 (4%)]\tLoss: -1949.739063\n",
            "Train Epoch: 44 [3840/60000 (6%)]\tLoss: -1969.520703\n",
            "Train Epoch: 44 [5120/60000 (9%)]\tLoss: -1953.378711\n",
            "Train Epoch: 44 [6400/60000 (11%)]\tLoss: -1932.050000\n",
            "Train Epoch: 44 [7680/60000 (13%)]\tLoss: -1959.087500\n",
            "Train Epoch: 44 [8960/60000 (15%)]\tLoss: -1975.314844\n",
            "Train Epoch: 44 [10240/60000 (17%)]\tLoss: -2196.312500\n",
            "Train Epoch: 44 [11520/60000 (19%)]\tLoss: -1951.958203\n",
            "Train Epoch: 44 [12800/60000 (21%)]\tLoss: -1988.039648\n",
            "Train Epoch: 44 [14080/60000 (23%)]\tLoss: -2148.983008\n",
            "Train Epoch: 44 [15360/60000 (26%)]\tLoss: -1947.121289\n",
            "Train Epoch: 44 [16640/60000 (28%)]\tLoss: -1951.271875\n",
            "Train Epoch: 44 [17920/60000 (30%)]\tLoss: -1950.059570\n",
            "Train Epoch: 44 [19200/60000 (32%)]\tLoss: -1993.168164\n",
            "Train Epoch: 44 [20480/60000 (34%)]\tLoss: -1942.254883\n",
            "Train Epoch: 44 [21760/60000 (36%)]\tLoss: -2028.109570\n",
            "Train Epoch: 44 [23040/60000 (38%)]\tLoss: -2021.638281\n",
            "Train Epoch: 44 [24320/60000 (41%)]\tLoss: -2031.725781\n",
            "Train Epoch: 44 [25600/60000 (43%)]\tLoss: -1918.947852\n",
            "Train Epoch: 44 [26880/60000 (45%)]\tLoss: -1953.625391\n",
            "Train Epoch: 44 [28160/60000 (47%)]\tLoss: -1997.733398\n",
            "Train Epoch: 44 [29440/60000 (49%)]\tLoss: -1933.950391\n",
            "Train Epoch: 44 [30720/60000 (51%)]\tLoss: -1962.285742\n",
            "Train Epoch: 44 [32000/60000 (53%)]\tLoss: -1985.578125\n",
            "Train Epoch: 44 [33280/60000 (55%)]\tLoss: -1923.996289\n",
            "Train Epoch: 44 [34560/60000 (58%)]\tLoss: -1915.492773\n",
            "Train Epoch: 44 [35840/60000 (60%)]\tLoss: -2016.073047\n",
            "Train Epoch: 44 [37120/60000 (62%)]\tLoss: -2065.409570\n",
            "Train Epoch: 44 [38400/60000 (64%)]\tLoss: -1960.924414\n",
            "Train Epoch: 44 [39680/60000 (66%)]\tLoss: -1947.290820\n",
            "Train Epoch: 44 [40960/60000 (68%)]\tLoss: -2013.771289\n",
            "Train Epoch: 44 [42240/60000 (70%)]\tLoss: -1947.825977\n",
            "Train Epoch: 44 [43520/60000 (72%)]\tLoss: -1938.862305\n",
            "Train Epoch: 44 [44800/60000 (75%)]\tLoss: -1982.715039\n",
            "Train Epoch: 44 [46080/60000 (77%)]\tLoss: -1981.229297\n",
            "Train Epoch: 44 [47360/60000 (79%)]\tLoss: -1956.665234\n",
            "Train Epoch: 44 [48640/60000 (81%)]\tLoss: -1946.706641\n",
            "Train Epoch: 44 [49920/60000 (83%)]\tLoss: -1966.952344\n",
            "Train Epoch: 44 [51200/60000 (85%)]\tLoss: -1986.275391\n",
            "Train Epoch: 44 [52480/60000 (87%)]\tLoss: -1947.693750\n",
            "Train Epoch: 44 [53760/60000 (90%)]\tLoss: -1958.683789\n",
            "Train Epoch: 44 [55040/60000 (92%)]\tLoss: -1913.375781\n",
            "Train Epoch: 44 [56320/60000 (94%)]\tLoss: -1946.614063\n",
            "Train Epoch: 44 [57600/60000 (96%)]\tLoss: -1963.164648\n",
            "Train Epoch: 44 [58880/60000 (98%)]\tLoss: -1913.110352\n",
            "====> Epoch: 44 Average loss: -19730.0654\n",
            "Train Epoch: 45 [0/60000 (0%)]\tLoss: -19368.324219\n",
            "Train Epoch: 45 [1280/60000 (2%)]\tLoss: -1993.698242\n",
            "Train Epoch: 45 [2560/60000 (4%)]\tLoss: -1924.526562\n",
            "Train Epoch: 45 [3840/60000 (6%)]\tLoss: -1974.063672\n",
            "Train Epoch: 45 [5120/60000 (9%)]\tLoss: -1942.438867\n",
            "Train Epoch: 45 [6400/60000 (11%)]\tLoss: -1930.671680\n",
            "Train Epoch: 45 [7680/60000 (13%)]\tLoss: -1951.144727\n",
            "Train Epoch: 45 [8960/60000 (15%)]\tLoss: -1965.395898\n",
            "Train Epoch: 45 [10240/60000 (17%)]\tLoss: -2190.593945\n",
            "Train Epoch: 45 [11520/60000 (19%)]\tLoss: -1950.367188\n",
            "Train Epoch: 45 [12800/60000 (21%)]\tLoss: -1976.375195\n",
            "Train Epoch: 45 [14080/60000 (23%)]\tLoss: -2070.449805\n",
            "Train Epoch: 45 [15360/60000 (26%)]\tLoss: -1945.437109\n",
            "Train Epoch: 45 [16640/60000 (28%)]\tLoss: -1949.176563\n",
            "Train Epoch: 45 [17920/60000 (30%)]\tLoss: -1939.077930\n",
            "Train Epoch: 45 [19200/60000 (32%)]\tLoss: -1998.675000\n",
            "Train Epoch: 45 [20480/60000 (34%)]\tLoss: -1922.319336\n",
            "Train Epoch: 45 [21760/60000 (36%)]\tLoss: -2007.998047\n",
            "Train Epoch: 45 [23040/60000 (38%)]\tLoss: -1995.758984\n",
            "Train Epoch: 45 [24320/60000 (41%)]\tLoss: -2029.999805\n",
            "Train Epoch: 45 [25600/60000 (43%)]\tLoss: -1907.418359\n",
            "Train Epoch: 45 [26880/60000 (45%)]\tLoss: -1973.169531\n",
            "Train Epoch: 45 [28160/60000 (47%)]\tLoss: -1981.676563\n",
            "Train Epoch: 45 [29440/60000 (49%)]\tLoss: -1933.219336\n",
            "Train Epoch: 45 [30720/60000 (51%)]\tLoss: -1943.083789\n",
            "Train Epoch: 45 [32000/60000 (53%)]\tLoss: -1947.743945\n",
            "Train Epoch: 45 [33280/60000 (55%)]\tLoss: -1922.360547\n",
            "Train Epoch: 45 [34560/60000 (58%)]\tLoss: -1908.107812\n",
            "Train Epoch: 45 [35840/60000 (60%)]\tLoss: -1972.248047\n",
            "Train Epoch: 45 [37120/60000 (62%)]\tLoss: -2077.609570\n",
            "Train Epoch: 45 [38400/60000 (64%)]\tLoss: -1965.993945\n",
            "Train Epoch: 45 [39680/60000 (66%)]\tLoss: -1947.601758\n",
            "Train Epoch: 45 [40960/60000 (68%)]\tLoss: -1993.703906\n",
            "Train Epoch: 45 [42240/60000 (70%)]\tLoss: -1944.893750\n",
            "Train Epoch: 45 [43520/60000 (72%)]\tLoss: -1936.117188\n",
            "Train Epoch: 45 [44800/60000 (75%)]\tLoss: -1973.143164\n",
            "Train Epoch: 45 [46080/60000 (77%)]\tLoss: -1979.211523\n",
            "Train Epoch: 45 [47360/60000 (79%)]\tLoss: -1967.565820\n",
            "Train Epoch: 45 [48640/60000 (81%)]\tLoss: -1963.753125\n",
            "Train Epoch: 45 [49920/60000 (83%)]\tLoss: -1988.888086\n",
            "Train Epoch: 45 [51200/60000 (85%)]\tLoss: -2023.692188\n",
            "Train Epoch: 45 [52480/60000 (87%)]\tLoss: -1946.587305\n",
            "Train Epoch: 45 [53760/60000 (90%)]\tLoss: -1962.444531\n",
            "Train Epoch: 45 [55040/60000 (92%)]\tLoss: -1911.283203\n",
            "Train Epoch: 45 [56320/60000 (94%)]\tLoss: -1948.255469\n",
            "Train Epoch: 45 [57600/60000 (96%)]\tLoss: -1954.665039\n",
            "Train Epoch: 45 [58880/60000 (98%)]\tLoss: -1912.738086\n",
            "====> Epoch: 45 Average loss: -19629.8939\n",
            "Train Epoch: 46 [0/60000 (0%)]\tLoss: -19359.013672\n",
            "Train Epoch: 46 [1280/60000 (2%)]\tLoss: -2001.733594\n",
            "Train Epoch: 46 [2560/60000 (4%)]\tLoss: -1924.134766\n",
            "Train Epoch: 46 [3840/60000 (6%)]\tLoss: -1956.983203\n",
            "Train Epoch: 46 [5120/60000 (9%)]\tLoss: -1956.209180\n",
            "Train Epoch: 46 [6400/60000 (11%)]\tLoss: -1928.541211\n",
            "Train Epoch: 46 [7680/60000 (13%)]\tLoss: -1950.578516\n",
            "Train Epoch: 46 [8960/60000 (15%)]\tLoss: -1979.497852\n",
            "Train Epoch: 46 [10240/60000 (17%)]\tLoss: -2169.462305\n",
            "Train Epoch: 46 [11520/60000 (19%)]\tLoss: -1952.717773\n",
            "Train Epoch: 46 [12800/60000 (21%)]\tLoss: -1946.410938\n",
            "Train Epoch: 46 [14080/60000 (23%)]\tLoss: -2076.299609\n",
            "Train Epoch: 46 [15360/60000 (26%)]\tLoss: -1948.255273\n",
            "Train Epoch: 46 [16640/60000 (28%)]\tLoss: -1950.862109\n",
            "Train Epoch: 46 [17920/60000 (30%)]\tLoss: -1954.047461\n",
            "Train Epoch: 46 [19200/60000 (32%)]\tLoss: -1998.269531\n",
            "Train Epoch: 46 [20480/60000 (34%)]\tLoss: -1917.377539\n",
            "Train Epoch: 46 [21760/60000 (36%)]\tLoss: -1992.088281\n",
            "Train Epoch: 46 [23040/60000 (38%)]\tLoss: -1986.924609\n",
            "Train Epoch: 46 [24320/60000 (41%)]\tLoss: -2018.186523\n",
            "Train Epoch: 46 [25600/60000 (43%)]\tLoss: -1918.051563\n",
            "Train Epoch: 46 [26880/60000 (45%)]\tLoss: -1964.239648\n",
            "Train Epoch: 46 [28160/60000 (47%)]\tLoss: -1971.838477\n",
            "Train Epoch: 46 [29440/60000 (49%)]\tLoss: -1928.738477\n",
            "Train Epoch: 46 [30720/60000 (51%)]\tLoss: -1941.866211\n",
            "Train Epoch: 46 [32000/60000 (53%)]\tLoss: -1964.149414\n",
            "Train Epoch: 46 [33280/60000 (55%)]\tLoss: -1919.039648\n",
            "Train Epoch: 46 [34560/60000 (58%)]\tLoss: -1908.293359\n",
            "Train Epoch: 46 [35840/60000 (60%)]\tLoss: -1982.246289\n",
            "Train Epoch: 46 [37120/60000 (62%)]\tLoss: -2069.666016\n",
            "Train Epoch: 46 [38400/60000 (64%)]\tLoss: -1959.872852\n",
            "Train Epoch: 46 [39680/60000 (66%)]\tLoss: -1953.614258\n",
            "Train Epoch: 46 [40960/60000 (68%)]\tLoss: -2009.160938\n",
            "Train Epoch: 46 [42240/60000 (70%)]\tLoss: -1956.482617\n",
            "Train Epoch: 46 [43520/60000 (72%)]\tLoss: -1934.594336\n",
            "Train Epoch: 46 [44800/60000 (75%)]\tLoss: -1976.982617\n",
            "Train Epoch: 46 [46080/60000 (77%)]\tLoss: -1974.911328\n",
            "Train Epoch: 46 [47360/60000 (79%)]\tLoss: -1975.638672\n",
            "Train Epoch: 46 [48640/60000 (81%)]\tLoss: -1981.309375\n",
            "Train Epoch: 46 [49920/60000 (83%)]\tLoss: -1990.619531\n",
            "Train Epoch: 46 [51200/60000 (85%)]\tLoss: -1977.838477\n",
            "Train Epoch: 46 [52480/60000 (87%)]\tLoss: -1947.167578\n",
            "Train Epoch: 46 [53760/60000 (90%)]\tLoss: -1961.768945\n",
            "Train Epoch: 46 [55040/60000 (92%)]\tLoss: -1908.991992\n",
            "Train Epoch: 46 [56320/60000 (94%)]\tLoss: -1954.201953\n",
            "Train Epoch: 46 [57600/60000 (96%)]\tLoss: -1951.208398\n",
            "Train Epoch: 46 [58880/60000 (98%)]\tLoss: -1911.324609\n",
            "====> Epoch: 46 Average loss: -19629.1112\n",
            "Train Epoch: 47 [0/60000 (0%)]\tLoss: -19367.164062\n",
            "Train Epoch: 47 [1280/60000 (2%)]\tLoss: -2000.221484\n",
            "Train Epoch: 47 [2560/60000 (4%)]\tLoss: -1924.206445\n",
            "Train Epoch: 47 [3840/60000 (6%)]\tLoss: -1969.011523\n",
            "Train Epoch: 47 [5120/60000 (9%)]\tLoss: -1965.940234\n",
            "Train Epoch: 47 [6400/60000 (11%)]\tLoss: -1936.725391\n",
            "Train Epoch: 47 [7680/60000 (13%)]\tLoss: -1957.749805\n",
            "Train Epoch: 47 [8960/60000 (15%)]\tLoss: -1965.673633\n",
            "Train Epoch: 47 [10240/60000 (17%)]\tLoss: -2145.117969\n",
            "Train Epoch: 47 [11520/60000 (19%)]\tLoss: -1954.815234\n",
            "Train Epoch: 47 [12800/60000 (21%)]\tLoss: -1993.540625\n",
            "Train Epoch: 47 [14080/60000 (23%)]\tLoss: -2079.821289\n",
            "Train Epoch: 47 [15360/60000 (26%)]\tLoss: -1949.710547\n",
            "Train Epoch: 47 [16640/60000 (28%)]\tLoss: -1949.233594\n",
            "Train Epoch: 47 [17920/60000 (30%)]\tLoss: -1947.382617\n",
            "Train Epoch: 47 [19200/60000 (32%)]\tLoss: -1998.281836\n",
            "Train Epoch: 47 [20480/60000 (34%)]\tLoss: -1921.025781\n",
            "Train Epoch: 47 [21760/60000 (36%)]\tLoss: -2020.552148\n",
            "Train Epoch: 47 [23040/60000 (38%)]\tLoss: -1995.710742\n",
            "Train Epoch: 47 [24320/60000 (41%)]\tLoss: -2026.724609\n",
            "Train Epoch: 47 [25600/60000 (43%)]\tLoss: -1916.511719\n",
            "Train Epoch: 47 [26880/60000 (45%)]\tLoss: -1971.680664\n",
            "Train Epoch: 47 [28160/60000 (47%)]\tLoss: -1976.883789\n",
            "Train Epoch: 47 [29440/60000 (49%)]\tLoss: -1939.286719\n",
            "Train Epoch: 47 [30720/60000 (51%)]\tLoss: -1940.598828\n",
            "Train Epoch: 47 [32000/60000 (53%)]\tLoss: -1959.164062\n",
            "Train Epoch: 47 [33280/60000 (55%)]\tLoss: -1917.860352\n",
            "Train Epoch: 47 [34560/60000 (58%)]\tLoss: -1912.104883\n",
            "Train Epoch: 47 [35840/60000 (60%)]\tLoss: -1970.539062\n",
            "Train Epoch: 47 [37120/60000 (62%)]\tLoss: -2061.254492\n",
            "Train Epoch: 47 [38400/60000 (64%)]\tLoss: -1966.036523\n",
            "Train Epoch: 47 [39680/60000 (66%)]\tLoss: -1958.440039\n",
            "Train Epoch: 47 [40960/60000 (68%)]\tLoss: -2013.506641\n",
            "Train Epoch: 47 [42240/60000 (70%)]\tLoss: -1967.068555\n",
            "Train Epoch: 47 [43520/60000 (72%)]\tLoss: -1933.697266\n",
            "Train Epoch: 47 [44800/60000 (75%)]\tLoss: -1964.697461\n",
            "Train Epoch: 47 [46080/60000 (77%)]\tLoss: -1972.731445\n",
            "Train Epoch: 47 [47360/60000 (79%)]\tLoss: -1958.787500\n",
            "Train Epoch: 47 [48640/60000 (81%)]\tLoss: -1963.322070\n",
            "Train Epoch: 47 [49920/60000 (83%)]\tLoss: -1999.101367\n",
            "Train Epoch: 47 [51200/60000 (85%)]\tLoss: -1990.660938\n",
            "Train Epoch: 47 [52480/60000 (87%)]\tLoss: -1960.413867\n",
            "Train Epoch: 47 [53760/60000 (90%)]\tLoss: -1954.281250\n",
            "Train Epoch: 47 [55040/60000 (92%)]\tLoss: -1910.087109\n",
            "Train Epoch: 47 [56320/60000 (94%)]\tLoss: -1964.406641\n",
            "Train Epoch: 47 [57600/60000 (96%)]\tLoss: -1948.071680\n",
            "Train Epoch: 47 [58880/60000 (98%)]\tLoss: -1912.312891\n",
            "====> Epoch: 47 Average loss: -19642.7747\n",
            "Train Epoch: 48 [0/60000 (0%)]\tLoss: -19529.541016\n",
            "Train Epoch: 48 [1280/60000 (2%)]\tLoss: -2017.410156\n",
            "Train Epoch: 48 [2560/60000 (4%)]\tLoss: -1949.903711\n",
            "Train Epoch: 48 [3840/60000 (6%)]\tLoss: -1961.568555\n",
            "Train Epoch: 48 [5120/60000 (9%)]\tLoss: -1960.284961\n",
            "Train Epoch: 48 [6400/60000 (11%)]\tLoss: -1926.941016\n",
            "Train Epoch: 48 [7680/60000 (13%)]\tLoss: -1939.170898\n",
            "Train Epoch: 48 [8960/60000 (15%)]\tLoss: -1983.358008\n",
            "Train Epoch: 48 [10240/60000 (17%)]\tLoss: -2179.368555\n",
            "Train Epoch: 48 [11520/60000 (19%)]\tLoss: -1953.817188\n",
            "Train Epoch: 48 [12800/60000 (21%)]\tLoss: -1970.762891\n",
            "Train Epoch: 48 [14080/60000 (23%)]\tLoss: -2127.933594\n",
            "Train Epoch: 48 [15360/60000 (26%)]\tLoss: -1951.151562\n",
            "Train Epoch: 48 [16640/60000 (28%)]\tLoss: -1948.790625\n",
            "Train Epoch: 48 [17920/60000 (30%)]\tLoss: -1949.875195\n",
            "Train Epoch: 48 [19200/60000 (32%)]\tLoss: -1988.829688\n",
            "Train Epoch: 48 [20480/60000 (34%)]\tLoss: -1929.601562\n",
            "Train Epoch: 48 [21760/60000 (36%)]\tLoss: -1978.372070\n",
            "Train Epoch: 48 [23040/60000 (38%)]\tLoss: -2034.333203\n",
            "Train Epoch: 48 [24320/60000 (41%)]\tLoss: -1988.222266\n",
            "Train Epoch: 48 [25600/60000 (43%)]\tLoss: -1922.662109\n",
            "Train Epoch: 48 [26880/60000 (45%)]\tLoss: -1948.180273\n",
            "Train Epoch: 48 [28160/60000 (47%)]\tLoss: -1967.716016\n",
            "Train Epoch: 48 [29440/60000 (49%)]\tLoss: -1945.029297\n",
            "Train Epoch: 48 [30720/60000 (51%)]\tLoss: -1954.003906\n",
            "Train Epoch: 48 [32000/60000 (53%)]\tLoss: -1948.797070\n",
            "Train Epoch: 48 [33280/60000 (55%)]\tLoss: -1914.739258\n",
            "Train Epoch: 48 [34560/60000 (58%)]\tLoss: -1911.589844\n",
            "Train Epoch: 48 [35840/60000 (60%)]\tLoss: -1983.670312\n",
            "Train Epoch: 48 [37120/60000 (62%)]\tLoss: -2070.495117\n",
            "Train Epoch: 48 [38400/60000 (64%)]\tLoss: -1959.967773\n",
            "Train Epoch: 48 [39680/60000 (66%)]\tLoss: -1957.910547\n",
            "Train Epoch: 48 [40960/60000 (68%)]\tLoss: -2002.144922\n",
            "Train Epoch: 48 [42240/60000 (70%)]\tLoss: -1974.777930\n",
            "Train Epoch: 48 [43520/60000 (72%)]\tLoss: -1935.732227\n",
            "Train Epoch: 48 [44800/60000 (75%)]\tLoss: -1966.239648\n",
            "Train Epoch: 48 [46080/60000 (77%)]\tLoss: -1988.011719\n",
            "Train Epoch: 48 [47360/60000 (79%)]\tLoss: -1998.322266\n",
            "Train Epoch: 48 [48640/60000 (81%)]\tLoss: -1988.802539\n",
            "Train Epoch: 48 [49920/60000 (83%)]\tLoss: -1993.100586\n",
            "Train Epoch: 48 [51200/60000 (85%)]\tLoss: -2009.227148\n",
            "Train Epoch: 48 [52480/60000 (87%)]\tLoss: -1952.728125\n",
            "Train Epoch: 48 [53760/60000 (90%)]\tLoss: -1954.917383\n",
            "Train Epoch: 48 [55040/60000 (92%)]\tLoss: -1908.648047\n",
            "Train Epoch: 48 [56320/60000 (94%)]\tLoss: -1959.288672\n",
            "Train Epoch: 48 [57600/60000 (96%)]\tLoss: -1947.103320\n",
            "Train Epoch: 48 [58880/60000 (98%)]\tLoss: -1913.789062\n",
            "====> Epoch: 48 Average loss: -19661.1363\n",
            "Train Epoch: 49 [0/60000 (0%)]\tLoss: -19479.726562\n",
            "Train Epoch: 49 [1280/60000 (2%)]\tLoss: -2062.906445\n",
            "Train Epoch: 49 [2560/60000 (4%)]\tLoss: -1941.487695\n",
            "Train Epoch: 49 [3840/60000 (6%)]\tLoss: -1964.622656\n",
            "Train Epoch: 49 [5120/60000 (9%)]\tLoss: -1960.608789\n",
            "Train Epoch: 49 [6400/60000 (11%)]\tLoss: -1922.754102\n",
            "Train Epoch: 49 [7680/60000 (13%)]\tLoss: -1990.729883\n",
            "Train Epoch: 49 [8960/60000 (15%)]\tLoss: -1977.432617\n",
            "Train Epoch: 49 [10240/60000 (17%)]\tLoss: -2183.814063\n",
            "Train Epoch: 49 [11520/60000 (19%)]\tLoss: -1951.031055\n",
            "Train Epoch: 49 [12800/60000 (21%)]\tLoss: -1969.317578\n",
            "Train Epoch: 49 [14080/60000 (23%)]\tLoss: -2106.189453\n",
            "Train Epoch: 49 [15360/60000 (26%)]\tLoss: -1947.255078\n",
            "Train Epoch: 49 [16640/60000 (28%)]\tLoss: -1946.209570\n",
            "Train Epoch: 49 [17920/60000 (30%)]\tLoss: -1947.592187\n",
            "Train Epoch: 49 [19200/60000 (32%)]\tLoss: -1998.764062\n",
            "Train Epoch: 49 [20480/60000 (34%)]\tLoss: -1920.872852\n",
            "Train Epoch: 49 [21760/60000 (36%)]\tLoss: -2031.138086\n",
            "Train Epoch: 49 [23040/60000 (38%)]\tLoss: -2002.215625\n",
            "Train Epoch: 49 [24320/60000 (41%)]\tLoss: -1997.539648\n",
            "Train Epoch: 49 [25600/60000 (43%)]\tLoss: -1924.540625\n",
            "Train Epoch: 49 [26880/60000 (45%)]\tLoss: -1949.599219\n",
            "Train Epoch: 49 [28160/60000 (47%)]\tLoss: -1981.168750\n",
            "Train Epoch: 49 [29440/60000 (49%)]\tLoss: -1949.957422\n",
            "Train Epoch: 49 [30720/60000 (51%)]\tLoss: -1950.246680\n",
            "Train Epoch: 49 [32000/60000 (53%)]\tLoss: -1987.444922\n",
            "Train Epoch: 49 [33280/60000 (55%)]\tLoss: -1919.354687\n",
            "Train Epoch: 49 [34560/60000 (58%)]\tLoss: -1907.551563\n",
            "Train Epoch: 49 [35840/60000 (60%)]\tLoss: -1988.744336\n",
            "Train Epoch: 49 [37120/60000 (62%)]\tLoss: -2062.412500\n",
            "Train Epoch: 49 [38400/60000 (64%)]\tLoss: -1946.435547\n",
            "Train Epoch: 49 [39680/60000 (66%)]\tLoss: -1958.767773\n",
            "Train Epoch: 49 [40960/60000 (68%)]\tLoss: -2023.627734\n",
            "Train Epoch: 49 [42240/60000 (70%)]\tLoss: -1965.148242\n",
            "Train Epoch: 49 [43520/60000 (72%)]\tLoss: -1932.471875\n",
            "Train Epoch: 49 [44800/60000 (75%)]\tLoss: -1999.252344\n",
            "Train Epoch: 49 [46080/60000 (77%)]\tLoss: -1997.441992\n",
            "Train Epoch: 49 [47360/60000 (79%)]\tLoss: -1965.677734\n",
            "Train Epoch: 49 [48640/60000 (81%)]\tLoss: -1975.280664\n",
            "Train Epoch: 49 [49920/60000 (83%)]\tLoss: -2004.605469\n",
            "Train Epoch: 49 [51200/60000 (85%)]\tLoss: -2004.607617\n",
            "Train Epoch: 49 [52480/60000 (87%)]\tLoss: -1958.852539\n",
            "Train Epoch: 49 [53760/60000 (90%)]\tLoss: -1957.627930\n",
            "Train Epoch: 49 [55040/60000 (92%)]\tLoss: -1907.699414\n",
            "Train Epoch: 49 [56320/60000 (94%)]\tLoss: -1948.741797\n",
            "Train Epoch: 49 [57600/60000 (96%)]\tLoss: -1952.658594\n",
            "Train Epoch: 49 [58880/60000 (98%)]\tLoss: -1910.723438\n",
            "====> Epoch: 49 Average loss: -19692.8230\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M0d3cRvRq1my",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "1aeb29ef-0008-4aa8-c5cd-13030665be98"
      },
      "source": [
        "from pathlib import Path\n",
        "import os\n",
        "\n",
        "!mkdir \"./cvae\"\n",
        "!mkdir \"./cvae/models\"\n",
        "\n",
        "def load_last_model():\n",
        "    models = Path('./cvae/models').rglob('*.pth')\n",
        "    \n",
        "    model_ids = [(int((f.as_posix()).split('_')[1]), f.as_posix()) for f in models]\n",
        "    start_epoch, last_cp = max(model_ids, key=lambda item:item[0])\n",
        "    print('Last checkpoint: ', last_cp)\n",
        "    model.load_state_dict(torch.load(last_cp))\n",
        "    return start_epoch, last_cp\n",
        "\n",
        "epochs = 50\n",
        "def resume_training():\n",
        "    start_epoch, _ = load_last_model()\n",
        "\n",
        "    for epoch in range(start_epoch + 1, start_epoch + epochs + 1):\n",
        "        train_loss = train(epoch)\n",
        "        # test_loss = test(epoch)\n",
        "        test_loss = 0\n",
        "        models = Path('./cvae/models').rglob('*.pth')\n",
        "        [os.remove(f.as_posix()) for f in models]\n",
        "        torch.save(model.state_dict(), './cvae/models/cvae-Epoch_{}_Train_loss_{:.4f}_Test_loss_{:.4f}.pth'.format(epoch, train_loss, test_loss))\n",
        "        # !cp -rf \"./vae-dfc\" \"drive/My Drive/Colab Notebooks/\"\n",
        "\n",
        "resume_training()\n",
        "\n",
        "# load_last_model()\n"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "mkdir: cannot create directory ‘./cvae’: File exists\n",
            "mkdir: cannot create directory ‘./cvae/models’: File exists\n",
            "Last checkpoint:  cvae/models/Epoch_50_Train_loss_0.0000_Test_loss_0.0000.pth\n",
            "Train Epoch: 51 [0/60000 (0%)]\tLoss: -19522.679688\n",
            "Train Epoch: 51 [1280/60000 (2%)]\tLoss: -2130.897070\n",
            "Train Epoch: 51 [2560/60000 (4%)]\tLoss: -1924.140625\n",
            "Train Epoch: 51 [3840/60000 (6%)]\tLoss: -1955.655078\n",
            "Train Epoch: 51 [5120/60000 (9%)]\tLoss: -1956.199805\n",
            "Train Epoch: 51 [6400/60000 (11%)]\tLoss: -1927.273828\n",
            "Train Epoch: 51 [7680/60000 (13%)]\tLoss: -1957.962695\n",
            "Train Epoch: 51 [8960/60000 (15%)]\tLoss: -1969.151172\n",
            "Train Epoch: 51 [10240/60000 (17%)]\tLoss: -2184.655469\n",
            "Train Epoch: 51 [11520/60000 (19%)]\tLoss: -1948.375391\n",
            "Train Epoch: 51 [12800/60000 (21%)]\tLoss: -1986.569336\n",
            "Train Epoch: 51 [14080/60000 (23%)]\tLoss: -2133.066211\n",
            "Train Epoch: 51 [15360/60000 (26%)]\tLoss: -1944.671484\n",
            "Train Epoch: 51 [16640/60000 (28%)]\tLoss: -1946.534375\n",
            "Train Epoch: 51 [17920/60000 (30%)]\tLoss: -1955.400195\n",
            "Train Epoch: 51 [19200/60000 (32%)]\tLoss: -2009.815625\n",
            "Train Epoch: 51 [20480/60000 (34%)]\tLoss: -1909.140820\n",
            "Train Epoch: 51 [21760/60000 (36%)]\tLoss: -2051.268555\n",
            "Train Epoch: 51 [23040/60000 (38%)]\tLoss: -2061.842773\n",
            "Train Epoch: 51 [24320/60000 (41%)]\tLoss: -2041.066602\n",
            "Train Epoch: 51 [25600/60000 (43%)]\tLoss: -1912.055078\n",
            "Train Epoch: 51 [26880/60000 (45%)]\tLoss: -1947.311523\n",
            "Train Epoch: 51 [28160/60000 (47%)]\tLoss: -1956.457422\n",
            "Train Epoch: 51 [29440/60000 (49%)]\tLoss: -1937.947461\n",
            "Train Epoch: 51 [30720/60000 (51%)]\tLoss: -1955.736914\n",
            "Train Epoch: 51 [32000/60000 (53%)]\tLoss: -1966.601367\n",
            "Train Epoch: 51 [33280/60000 (55%)]\tLoss: -1918.998437\n",
            "Train Epoch: 51 [34560/60000 (58%)]\tLoss: -1908.508008\n",
            "Train Epoch: 51 [35840/60000 (60%)]\tLoss: -1995.109375\n",
            "Train Epoch: 51 [37120/60000 (62%)]\tLoss: -2052.873437\n",
            "Train Epoch: 51 [38400/60000 (64%)]\tLoss: -1952.816211\n",
            "Train Epoch: 51 [39680/60000 (66%)]\tLoss: -1957.769336\n",
            "Train Epoch: 51 [40960/60000 (68%)]\tLoss: -1983.575586\n",
            "Train Epoch: 51 [42240/60000 (70%)]\tLoss: -1948.884961\n",
            "Train Epoch: 51 [43520/60000 (72%)]\tLoss: -1930.570898\n",
            "Train Epoch: 51 [44800/60000 (75%)]\tLoss: -1969.854297\n",
            "Train Epoch: 51 [46080/60000 (77%)]\tLoss: -1976.538672\n",
            "Train Epoch: 51 [47360/60000 (79%)]\tLoss: -1964.021289\n",
            "Train Epoch: 51 [48640/60000 (81%)]\tLoss: -2000.450586\n",
            "Train Epoch: 51 [49920/60000 (83%)]\tLoss: -2000.597266\n",
            "Train Epoch: 51 [51200/60000 (85%)]\tLoss: -2020.145117\n",
            "Train Epoch: 51 [52480/60000 (87%)]\tLoss: -1950.332031\n",
            "Train Epoch: 51 [53760/60000 (90%)]\tLoss: -1956.773633\n",
            "Train Epoch: 51 [55040/60000 (92%)]\tLoss: -1922.267188\n",
            "Train Epoch: 51 [56320/60000 (94%)]\tLoss: -1946.119922\n",
            "Train Epoch: 51 [57600/60000 (96%)]\tLoss: -1951.517969\n",
            "Train Epoch: 51 [58880/60000 (98%)]\tLoss: -1909.183008\n",
            "====> Epoch: 51 Average loss: -19711.2899\n",
            "Train Epoch: 52 [0/60000 (0%)]\tLoss: -19579.867188\n",
            "Train Epoch: 52 [1280/60000 (2%)]\tLoss: -2160.324609\n",
            "Train Epoch: 52 [2560/60000 (4%)]\tLoss: -1952.635352\n",
            "Train Epoch: 52 [3840/60000 (6%)]\tLoss: -1960.655859\n",
            "Train Epoch: 52 [5120/60000 (9%)]\tLoss: -2000.059180\n",
            "Train Epoch: 52 [6400/60000 (11%)]\tLoss: -1923.780664\n",
            "Train Epoch: 52 [7680/60000 (13%)]\tLoss: -1952.934570\n",
            "Train Epoch: 52 [8960/60000 (15%)]\tLoss: -1985.685156\n",
            "Train Epoch: 52 [10240/60000 (17%)]\tLoss: -2235.394141\n",
            "Train Epoch: 52 [11520/60000 (19%)]\tLoss: -1949.590430\n",
            "Train Epoch: 52 [12800/60000 (21%)]\tLoss: -1987.260742\n",
            "Train Epoch: 52 [14080/60000 (23%)]\tLoss: -2138.723828\n",
            "Train Epoch: 52 [15360/60000 (26%)]\tLoss: -1954.928320\n",
            "Train Epoch: 52 [16640/60000 (28%)]\tLoss: -1956.136914\n",
            "Train Epoch: 52 [17920/60000 (30%)]\tLoss: -1944.655664\n",
            "Train Epoch: 52 [19200/60000 (32%)]\tLoss: -2021.450586\n",
            "Train Epoch: 52 [20480/60000 (34%)]\tLoss: -1930.748828\n",
            "Train Epoch: 52 [21760/60000 (36%)]\tLoss: -2035.514258\n",
            "Train Epoch: 52 [23040/60000 (38%)]\tLoss: -2048.706641\n",
            "Train Epoch: 52 [24320/60000 (41%)]\tLoss: -2038.756836\n",
            "Train Epoch: 52 [25600/60000 (43%)]\tLoss: -1929.298828\n",
            "Train Epoch: 52 [26880/60000 (45%)]\tLoss: -1949.271094\n",
            "Train Epoch: 52 [28160/60000 (47%)]\tLoss: -1979.124023\n",
            "Train Epoch: 52 [29440/60000 (49%)]\tLoss: -1935.030469\n",
            "Train Epoch: 52 [30720/60000 (51%)]\tLoss: -1949.812109\n",
            "Train Epoch: 52 [32000/60000 (53%)]\tLoss: -1969.770313\n",
            "Train Epoch: 52 [33280/60000 (55%)]\tLoss: -1915.966797\n",
            "Train Epoch: 52 [34560/60000 (58%)]\tLoss: -1906.178125\n",
            "Train Epoch: 52 [35840/60000 (60%)]\tLoss: -1996.774023\n",
            "Train Epoch: 52 [37120/60000 (62%)]\tLoss: -2054.058008\n",
            "Train Epoch: 52 [38400/60000 (64%)]\tLoss: -1959.718945\n",
            "Train Epoch: 52 [39680/60000 (66%)]\tLoss: -1981.276758\n",
            "Train Epoch: 52 [40960/60000 (68%)]\tLoss: -2019.999219\n",
            "Train Epoch: 52 [42240/60000 (70%)]\tLoss: -1978.905859\n",
            "Train Epoch: 52 [43520/60000 (72%)]\tLoss: -1931.635937\n",
            "Train Epoch: 52 [44800/60000 (75%)]\tLoss: -1996.332813\n",
            "Train Epoch: 52 [46080/60000 (77%)]\tLoss: -2019.873437\n",
            "Train Epoch: 52 [47360/60000 (79%)]\tLoss: -1971.120703\n",
            "Train Epoch: 52 [48640/60000 (81%)]\tLoss: -1987.598633\n",
            "Train Epoch: 52 [49920/60000 (83%)]\tLoss: -1997.731250\n",
            "Train Epoch: 52 [51200/60000 (85%)]\tLoss: -2013.091016\n",
            "Train Epoch: 52 [52480/60000 (87%)]\tLoss: -1957.636719\n",
            "Train Epoch: 52 [53760/60000 (90%)]\tLoss: -1967.393750\n",
            "Train Epoch: 52 [55040/60000 (92%)]\tLoss: -1918.716406\n",
            "Train Epoch: 52 [56320/60000 (94%)]\tLoss: -1961.968555\n",
            "Train Epoch: 52 [57600/60000 (96%)]\tLoss: -1950.003125\n",
            "Train Epoch: 52 [58880/60000 (98%)]\tLoss: -1910.613281\n",
            "====> Epoch: 52 Average loss: -19773.7370\n",
            "Train Epoch: 53 [0/60000 (0%)]\tLoss: -19511.748047\n",
            "Train Epoch: 53 [1280/60000 (2%)]\tLoss: -2030.750195\n",
            "Train Epoch: 53 [2560/60000 (4%)]\tLoss: -1948.262500\n",
            "Train Epoch: 53 [3840/60000 (6%)]\tLoss: -1972.902734\n",
            "Train Epoch: 53 [5120/60000 (9%)]\tLoss: -2011.524023\n",
            "Train Epoch: 53 [6400/60000 (11%)]\tLoss: -1934.178516\n",
            "Train Epoch: 53 [7680/60000 (13%)]\tLoss: -1952.500781\n",
            "Train Epoch: 53 [8960/60000 (15%)]\tLoss: -1972.273047\n",
            "Train Epoch: 53 [10240/60000 (17%)]\tLoss: -2211.086914\n",
            "Train Epoch: 53 [11520/60000 (19%)]\tLoss: -1939.020898\n",
            "Train Epoch: 53 [12800/60000 (21%)]\tLoss: -1974.300781\n",
            "Train Epoch: 53 [14080/60000 (23%)]\tLoss: -2185.469922\n",
            "Train Epoch: 53 [15360/60000 (26%)]\tLoss: -1947.274414\n",
            "Train Epoch: 53 [16640/60000 (28%)]\tLoss: -1945.454883\n",
            "Train Epoch: 53 [17920/60000 (30%)]\tLoss: -1944.644141\n",
            "Train Epoch: 53 [19200/60000 (32%)]\tLoss: -2038.884766\n",
            "Train Epoch: 53 [20480/60000 (34%)]\tLoss: -1920.683984\n",
            "Train Epoch: 53 [21760/60000 (36%)]\tLoss: -2091.037695\n",
            "Train Epoch: 53 [23040/60000 (38%)]\tLoss: -1992.807031\n",
            "Train Epoch: 53 [24320/60000 (41%)]\tLoss: -2033.053711\n",
            "Train Epoch: 53 [25600/60000 (43%)]\tLoss: -1907.810352\n",
            "Train Epoch: 53 [26880/60000 (45%)]\tLoss: -1955.338477\n",
            "Train Epoch: 53 [28160/60000 (47%)]\tLoss: -1966.681836\n",
            "Train Epoch: 53 [29440/60000 (49%)]\tLoss: -1942.350391\n",
            "Train Epoch: 53 [30720/60000 (51%)]\tLoss: -1937.366406\n",
            "Train Epoch: 53 [32000/60000 (53%)]\tLoss: -1986.247852\n",
            "Train Epoch: 53 [33280/60000 (55%)]\tLoss: -1917.634375\n",
            "Train Epoch: 53 [34560/60000 (58%)]\tLoss: -1903.914062\n",
            "Train Epoch: 53 [35840/60000 (60%)]\tLoss: -1999.774219\n",
            "Train Epoch: 53 [37120/60000 (62%)]\tLoss: -2046.477539\n",
            "Train Epoch: 53 [38400/60000 (64%)]\tLoss: -1969.906250\n",
            "Train Epoch: 53 [39680/60000 (66%)]\tLoss: -1947.577930\n",
            "Train Epoch: 53 [40960/60000 (68%)]\tLoss: -2051.950000\n",
            "Train Epoch: 53 [42240/60000 (70%)]\tLoss: -1954.104102\n",
            "Train Epoch: 53 [43520/60000 (72%)]\tLoss: -1935.559766\n",
            "Train Epoch: 53 [44800/60000 (75%)]\tLoss: -1993.226562\n",
            "Train Epoch: 53 [46080/60000 (77%)]\tLoss: -2013.248242\n",
            "Train Epoch: 53 [47360/60000 (79%)]\tLoss: -1964.216016\n",
            "Train Epoch: 53 [48640/60000 (81%)]\tLoss: -1991.415039\n",
            "Train Epoch: 53 [49920/60000 (83%)]\tLoss: -2005.009180\n",
            "Train Epoch: 53 [51200/60000 (85%)]\tLoss: -2039.582227\n",
            "Train Epoch: 53 [52480/60000 (87%)]\tLoss: -1946.784180\n",
            "Train Epoch: 53 [53760/60000 (90%)]\tLoss: -1953.752539\n",
            "Train Epoch: 53 [55040/60000 (92%)]\tLoss: -1913.832227\n",
            "Train Epoch: 53 [56320/60000 (94%)]\tLoss: -1956.621289\n",
            "Train Epoch: 53 [57600/60000 (96%)]\tLoss: -1964.629297\n",
            "Train Epoch: 53 [58880/60000 (98%)]\tLoss: -1908.316211\n",
            "====> Epoch: 53 Average loss: -19771.6545\n",
            "Train Epoch: 54 [0/60000 (0%)]\tLoss: -19467.414062\n",
            "Train Epoch: 54 [1280/60000 (2%)]\tLoss: -2071.371094\n",
            "Train Epoch: 54 [2560/60000 (4%)]\tLoss: -1942.797266\n",
            "Train Epoch: 54 [3840/60000 (6%)]\tLoss: -1968.414062\n",
            "Train Epoch: 54 [5120/60000 (9%)]\tLoss: -1985.075586\n",
            "Train Epoch: 54 [6400/60000 (11%)]\tLoss: -1945.398047\n",
            "Train Epoch: 54 [7680/60000 (13%)]\tLoss: -1969.600391\n",
            "Train Epoch: 54 [8960/60000 (15%)]\tLoss: -1993.851758\n",
            "Train Epoch: 54 [10240/60000 (17%)]\tLoss: -2223.765430\n",
            "Train Epoch: 54 [11520/60000 (19%)]\tLoss: -1944.116406\n",
            "Train Epoch: 54 [12800/60000 (21%)]\tLoss: -1991.411914\n",
            "Train Epoch: 54 [14080/60000 (23%)]\tLoss: -2199.465430\n",
            "Train Epoch: 54 [15360/60000 (26%)]\tLoss: -1955.929102\n",
            "Train Epoch: 54 [16640/60000 (28%)]\tLoss: -1943.792969\n",
            "Train Epoch: 54 [17920/60000 (30%)]\tLoss: -1954.130273\n",
            "Train Epoch: 54 [19200/60000 (32%)]\tLoss: -2014.687305\n",
            "Train Epoch: 54 [20480/60000 (34%)]\tLoss: -1920.073633\n",
            "Train Epoch: 54 [21760/60000 (36%)]\tLoss: -2116.674609\n",
            "Train Epoch: 54 [23040/60000 (38%)]\tLoss: -2018.663672\n",
            "Train Epoch: 54 [24320/60000 (41%)]\tLoss: -2033.748828\n",
            "Train Epoch: 54 [25600/60000 (43%)]\tLoss: -1906.724219\n",
            "Train Epoch: 54 [26880/60000 (45%)]\tLoss: -1962.297266\n",
            "Train Epoch: 54 [28160/60000 (47%)]\tLoss: -1998.676172\n",
            "Train Epoch: 54 [29440/60000 (49%)]\tLoss: -1942.359180\n",
            "Train Epoch: 54 [30720/60000 (51%)]\tLoss: -1939.580859\n",
            "Train Epoch: 54 [32000/60000 (53%)]\tLoss: -1988.449609\n",
            "Train Epoch: 54 [33280/60000 (55%)]\tLoss: -1918.567773\n",
            "Train Epoch: 54 [34560/60000 (58%)]\tLoss: -1905.075586\n",
            "Train Epoch: 54 [35840/60000 (60%)]\tLoss: -1998.380859\n",
            "Train Epoch: 54 [37120/60000 (62%)]\tLoss: -2093.386914\n",
            "Train Epoch: 54 [38400/60000 (64%)]\tLoss: -1975.915039\n",
            "Train Epoch: 54 [39680/60000 (66%)]\tLoss: -1956.606055\n",
            "Train Epoch: 54 [40960/60000 (68%)]\tLoss: -2010.515820\n",
            "Train Epoch: 54 [42240/60000 (70%)]\tLoss: -1977.889453\n",
            "Train Epoch: 54 [43520/60000 (72%)]\tLoss: -1964.583594\n",
            "Train Epoch: 54 [44800/60000 (75%)]\tLoss: -1970.739453\n",
            "Train Epoch: 54 [46080/60000 (77%)]\tLoss: -1996.716797\n",
            "Train Epoch: 54 [47360/60000 (79%)]\tLoss: -1986.394531\n",
            "Train Epoch: 54 [48640/60000 (81%)]\tLoss: -2001.832422\n",
            "Train Epoch: 54 [49920/60000 (83%)]\tLoss: -2006.399609\n",
            "Train Epoch: 54 [51200/60000 (85%)]\tLoss: -2029.645703\n",
            "Train Epoch: 54 [52480/60000 (87%)]\tLoss: -1967.936133\n",
            "Train Epoch: 54 [53760/60000 (90%)]\tLoss: -1957.286328\n",
            "Train Epoch: 54 [55040/60000 (92%)]\tLoss: -1911.595117\n",
            "Train Epoch: 54 [56320/60000 (94%)]\tLoss: -1959.575195\n",
            "Train Epoch: 54 [57600/60000 (96%)]\tLoss: -1968.582813\n",
            "Train Epoch: 54 [58880/60000 (98%)]\tLoss: -1909.187109\n",
            "====> Epoch: 54 Average loss: -19794.6948\n",
            "Train Epoch: 55 [0/60000 (0%)]\tLoss: -19481.912109\n",
            "Train Epoch: 55 [1280/60000 (2%)]\tLoss: -2102.098633\n",
            "Train Epoch: 55 [2560/60000 (4%)]\tLoss: -1930.833398\n",
            "Train Epoch: 55 [3840/60000 (6%)]\tLoss: -1967.078516\n",
            "Train Epoch: 55 [5120/60000 (9%)]\tLoss: -1991.133594\n",
            "Train Epoch: 55 [6400/60000 (11%)]\tLoss: -1934.852734\n",
            "Train Epoch: 55 [7680/60000 (13%)]\tLoss: -1955.769727\n",
            "Train Epoch: 55 [8960/60000 (15%)]\tLoss: -1978.306836\n",
            "Train Epoch: 55 [10240/60000 (17%)]\tLoss: -2245.866211\n",
            "Train Epoch: 55 [11520/60000 (19%)]\tLoss: -1948.792773\n",
            "Train Epoch: 55 [12800/60000 (21%)]\tLoss: -1992.318945\n",
            "Train Epoch: 55 [14080/60000 (23%)]\tLoss: -2196.937695\n",
            "Train Epoch: 55 [15360/60000 (26%)]\tLoss: -1948.411328\n",
            "Train Epoch: 55 [16640/60000 (28%)]\tLoss: -1942.405078\n",
            "Train Epoch: 55 [17920/60000 (30%)]\tLoss: -1954.158789\n",
            "Train Epoch: 55 [19200/60000 (32%)]\tLoss: -2022.962891\n",
            "Train Epoch: 55 [20480/60000 (34%)]\tLoss: -1925.283594\n",
            "Train Epoch: 55 [21760/60000 (36%)]\tLoss: -2122.991016\n",
            "Train Epoch: 55 [23040/60000 (38%)]\tLoss: -2038.193750\n",
            "Train Epoch: 55 [24320/60000 (41%)]\tLoss: -2026.679297\n",
            "Train Epoch: 55 [25600/60000 (43%)]\tLoss: -1913.103516\n",
            "Train Epoch: 55 [26880/60000 (45%)]\tLoss: -1947.972461\n",
            "Train Epoch: 55 [28160/60000 (47%)]\tLoss: -2034.990820\n",
            "Train Epoch: 55 [29440/60000 (49%)]\tLoss: -1940.346680\n",
            "Train Epoch: 55 [30720/60000 (51%)]\tLoss: -1965.979297\n",
            "Train Epoch: 55 [32000/60000 (53%)]\tLoss: -1976.073828\n",
            "Train Epoch: 55 [33280/60000 (55%)]\tLoss: -1919.469336\n",
            "Train Epoch: 55 [34560/60000 (58%)]\tLoss: -1912.958008\n",
            "Train Epoch: 55 [35840/60000 (60%)]\tLoss: -1991.526758\n",
            "Train Epoch: 55 [37120/60000 (62%)]\tLoss: -2094.823242\n",
            "Train Epoch: 55 [38400/60000 (64%)]\tLoss: -1968.836719\n",
            "Train Epoch: 55 [39680/60000 (66%)]\tLoss: -1968.176172\n",
            "Train Epoch: 55 [40960/60000 (68%)]\tLoss: -2006.420117\n",
            "Train Epoch: 55 [42240/60000 (70%)]\tLoss: -1982.476953\n",
            "Train Epoch: 55 [43520/60000 (72%)]\tLoss: -1987.705469\n",
            "Train Epoch: 55 [44800/60000 (75%)]\tLoss: -1997.587500\n",
            "Train Epoch: 55 [46080/60000 (77%)]\tLoss: -2003.738086\n",
            "Train Epoch: 55 [47360/60000 (79%)]\tLoss: -2015.017578\n",
            "Train Epoch: 55 [48640/60000 (81%)]\tLoss: -1984.007617\n",
            "Train Epoch: 55 [49920/60000 (83%)]\tLoss: -2001.029102\n",
            "Train Epoch: 55 [51200/60000 (85%)]\tLoss: -2035.600391\n",
            "Train Epoch: 55 [52480/60000 (87%)]\tLoss: -1944.498047\n",
            "Train Epoch: 55 [53760/60000 (90%)]\tLoss: -1968.841602\n",
            "Train Epoch: 55 [55040/60000 (92%)]\tLoss: -1915.688672\n",
            "Train Epoch: 55 [56320/60000 (94%)]\tLoss: -1945.062109\n",
            "Train Epoch: 55 [57600/60000 (96%)]\tLoss: -1956.251563\n",
            "Train Epoch: 55 [58880/60000 (98%)]\tLoss: -1924.367773\n",
            "====> Epoch: 55 Average loss: -19831.7321\n",
            "Train Epoch: 56 [0/60000 (0%)]\tLoss: -19717.351562\n",
            "Train Epoch: 56 [1280/60000 (2%)]\tLoss: -2157.467773\n",
            "Train Epoch: 56 [2560/60000 (4%)]\tLoss: -1928.001563\n",
            "Train Epoch: 56 [3840/60000 (6%)]\tLoss: -1985.971875\n",
            "Train Epoch: 56 [5120/60000 (9%)]\tLoss: -2025.389062\n",
            "Train Epoch: 56 [6400/60000 (11%)]\tLoss: -1948.798242\n",
            "Train Epoch: 56 [7680/60000 (13%)]\tLoss: -1954.191211\n",
            "Train Epoch: 56 [8960/60000 (15%)]\tLoss: -2007.579297\n",
            "Train Epoch: 56 [10240/60000 (17%)]\tLoss: -2256.172852\n",
            "Train Epoch: 56 [11520/60000 (19%)]\tLoss: -1947.899609\n",
            "Train Epoch: 56 [12800/60000 (21%)]\tLoss: -1998.626758\n",
            "Train Epoch: 56 [14080/60000 (23%)]\tLoss: -2222.844336\n",
            "Train Epoch: 56 [15360/60000 (26%)]\tLoss: -1969.930469\n",
            "Train Epoch: 56 [16640/60000 (28%)]\tLoss: -1945.457031\n",
            "Train Epoch: 56 [17920/60000 (30%)]\tLoss: -1956.208203\n",
            "Train Epoch: 56 [19200/60000 (32%)]\tLoss: -2050.976758\n",
            "Train Epoch: 56 [20480/60000 (34%)]\tLoss: -1928.506250\n",
            "Train Epoch: 56 [21760/60000 (36%)]\tLoss: -2085.594141\n",
            "Train Epoch: 56 [23040/60000 (38%)]\tLoss: -2053.214258\n",
            "Train Epoch: 56 [24320/60000 (41%)]\tLoss: -2034.438672\n",
            "Train Epoch: 56 [25600/60000 (43%)]\tLoss: -1921.868750\n",
            "Train Epoch: 56 [26880/60000 (45%)]\tLoss: -1953.957617\n",
            "Train Epoch: 56 [28160/60000 (47%)]\tLoss: -2037.251367\n",
            "Train Epoch: 56 [29440/60000 (49%)]\tLoss: -1936.361914\n",
            "Train Epoch: 56 [30720/60000 (51%)]\tLoss: -1962.421094\n",
            "Train Epoch: 56 [32000/60000 (53%)]\tLoss: -2003.217578\n",
            "Train Epoch: 56 [33280/60000 (55%)]\tLoss: -1920.742969\n",
            "Train Epoch: 56 [34560/60000 (58%)]\tLoss: -1958.455664\n",
            "Train Epoch: 56 [35840/60000 (60%)]\tLoss: -1986.826562\n",
            "Train Epoch: 56 [37120/60000 (62%)]\tLoss: -2129.155664\n",
            "Train Epoch: 56 [38400/60000 (64%)]\tLoss: -1987.055664\n",
            "Train Epoch: 56 [39680/60000 (66%)]\tLoss: -1967.683398\n",
            "Train Epoch: 56 [40960/60000 (68%)]\tLoss: -2012.197461\n",
            "Train Epoch: 56 [42240/60000 (70%)]\tLoss: -1966.433203\n",
            "Train Epoch: 56 [43520/60000 (72%)]\tLoss: -1957.226367\n",
            "Train Epoch: 56 [44800/60000 (75%)]\tLoss: -1995.165820\n",
            "Train Epoch: 56 [46080/60000 (77%)]\tLoss: -2022.525195\n",
            "Train Epoch: 56 [47360/60000 (79%)]\tLoss: -2061.811914\n",
            "Train Epoch: 56 [48640/60000 (81%)]\tLoss: -2007.339844\n",
            "Train Epoch: 56 [49920/60000 (83%)]\tLoss: -2028.104102\n",
            "Train Epoch: 56 [51200/60000 (85%)]\tLoss: -2075.538672\n",
            "Train Epoch: 56 [52480/60000 (87%)]\tLoss: -1949.954492\n",
            "Train Epoch: 56 [53760/60000 (90%)]\tLoss: -1963.129297\n",
            "Train Epoch: 56 [55040/60000 (92%)]\tLoss: -1907.616992\n",
            "Train Epoch: 56 [56320/60000 (94%)]\tLoss: -1939.713672\n",
            "Train Epoch: 56 [57600/60000 (96%)]\tLoss: -1949.494531\n",
            "Train Epoch: 56 [58880/60000 (98%)]\tLoss: -1914.981055\n",
            "====> Epoch: 56 Average loss: -19898.9678\n",
            "Train Epoch: 57 [0/60000 (0%)]\tLoss: -19715.031250\n",
            "Train Epoch: 57 [1280/60000 (2%)]\tLoss: -2166.810156\n",
            "Train Epoch: 57 [2560/60000 (4%)]\tLoss: -1945.554688\n",
            "Train Epoch: 57 [3840/60000 (6%)]\tLoss: -1989.820703\n",
            "Train Epoch: 57 [5120/60000 (9%)]\tLoss: -2003.208594\n",
            "Train Epoch: 57 [6400/60000 (11%)]\tLoss: -1944.408203\n",
            "Train Epoch: 57 [7680/60000 (13%)]\tLoss: -1961.675195\n",
            "Train Epoch: 57 [8960/60000 (15%)]\tLoss: -1996.859961\n",
            "Train Epoch: 57 [10240/60000 (17%)]\tLoss: -2193.006250\n",
            "Train Epoch: 57 [11520/60000 (19%)]\tLoss: -1946.263281\n",
            "Train Epoch: 57 [12800/60000 (21%)]\tLoss: -1981.455273\n",
            "Train Epoch: 57 [14080/60000 (23%)]\tLoss: -2212.846289\n",
            "Train Epoch: 57 [15360/60000 (26%)]\tLoss: -1989.030859\n",
            "Train Epoch: 57 [16640/60000 (28%)]\tLoss: -1946.740234\n",
            "Train Epoch: 57 [17920/60000 (30%)]\tLoss: -1936.242578\n",
            "Train Epoch: 57 [19200/60000 (32%)]\tLoss: -2036.323437\n",
            "Train Epoch: 57 [20480/60000 (34%)]\tLoss: -1925.442383\n",
            "Train Epoch: 57 [21760/60000 (36%)]\tLoss: -2183.517578\n",
            "Train Epoch: 57 [23040/60000 (38%)]\tLoss: -2057.923633\n",
            "Train Epoch: 57 [24320/60000 (41%)]\tLoss: -2029.180664\n",
            "Train Epoch: 57 [25600/60000 (43%)]\tLoss: -1937.183398\n",
            "Train Epoch: 57 [26880/60000 (45%)]\tLoss: -1962.515820\n",
            "Train Epoch: 57 [28160/60000 (47%)]\tLoss: -2045.711719\n",
            "Train Epoch: 57 [29440/60000 (49%)]\tLoss: -1931.630273\n",
            "Train Epoch: 57 [30720/60000 (51%)]\tLoss: -1941.264648\n",
            "Train Epoch: 57 [32000/60000 (53%)]\tLoss: -1992.681055\n",
            "Train Epoch: 57 [33280/60000 (55%)]\tLoss: -1955.986914\n",
            "Train Epoch: 57 [34560/60000 (58%)]\tLoss: -2089.815820\n",
            "Train Epoch: 57 [35840/60000 (60%)]\tLoss: -2004.362500\n",
            "Train Epoch: 57 [37120/60000 (62%)]\tLoss: -2063.449023\n",
            "Train Epoch: 57 [38400/60000 (64%)]\tLoss: -1975.042969\n",
            "Train Epoch: 57 [39680/60000 (66%)]\tLoss: -1981.088086\n",
            "Train Epoch: 57 [40960/60000 (68%)]\tLoss: -2033.398242\n",
            "Train Epoch: 57 [42240/60000 (70%)]\tLoss: -1974.429688\n",
            "Train Epoch: 57 [43520/60000 (72%)]\tLoss: -1975.712891\n",
            "Train Epoch: 57 [44800/60000 (75%)]\tLoss: -2031.417383\n",
            "Train Epoch: 57 [46080/60000 (77%)]\tLoss: -2045.477148\n",
            "Train Epoch: 57 [47360/60000 (79%)]\tLoss: -2067.676953\n",
            "Train Epoch: 57 [48640/60000 (81%)]\tLoss: -2003.162109\n",
            "Train Epoch: 57 [49920/60000 (83%)]\tLoss: -2014.872266\n",
            "Train Epoch: 57 [51200/60000 (85%)]\tLoss: -2127.701758\n",
            "Train Epoch: 57 [52480/60000 (87%)]\tLoss: -1968.713867\n",
            "Train Epoch: 57 [53760/60000 (90%)]\tLoss: -1968.522461\n",
            "Train Epoch: 57 [55040/60000 (92%)]\tLoss: -1928.693945\n",
            "Train Epoch: 57 [56320/60000 (94%)]\tLoss: -1940.417773\n",
            "Train Epoch: 57 [57600/60000 (96%)]\tLoss: -1970.772852\n",
            "Train Epoch: 57 [58880/60000 (98%)]\tLoss: -1912.610547\n",
            "====> Epoch: 57 Average loss: -19970.3709\n",
            "Train Epoch: 58 [0/60000 (0%)]\tLoss: -19638.001953\n",
            "Train Epoch: 58 [1280/60000 (2%)]\tLoss: -2243.748633\n",
            "Train Epoch: 58 [2560/60000 (4%)]\tLoss: -1933.051758\n",
            "Train Epoch: 58 [3840/60000 (6%)]\tLoss: -1959.098828\n",
            "Train Epoch: 58 [5120/60000 (9%)]\tLoss: -2004.430469\n",
            "Train Epoch: 58 [6400/60000 (11%)]\tLoss: -1959.663086\n",
            "Train Epoch: 58 [7680/60000 (13%)]\tLoss: -1971.279687\n",
            "Train Epoch: 58 [8960/60000 (15%)]\tLoss: -1965.767188\n",
            "Train Epoch: 58 [10240/60000 (17%)]\tLoss: -2243.221289\n",
            "Train Epoch: 58 [11520/60000 (19%)]\tLoss: -1947.041602\n",
            "Train Epoch: 58 [12800/60000 (21%)]\tLoss: -1979.244141\n",
            "Train Epoch: 58 [14080/60000 (23%)]\tLoss: -2178.484180\n",
            "Train Epoch: 58 [15360/60000 (26%)]\tLoss: -2004.765625\n",
            "Train Epoch: 58 [16640/60000 (28%)]\tLoss: -1947.714062\n",
            "Train Epoch: 58 [17920/60000 (30%)]\tLoss: -1952.030664\n",
            "Train Epoch: 58 [19200/60000 (32%)]\tLoss: -2024.451367\n",
            "Train Epoch: 58 [20480/60000 (34%)]\tLoss: -1923.372656\n",
            "Train Epoch: 58 [21760/60000 (36%)]\tLoss: -2094.139258\n",
            "Train Epoch: 58 [23040/60000 (38%)]\tLoss: -2108.866992\n",
            "Train Epoch: 58 [24320/60000 (41%)]\tLoss: -2031.362695\n",
            "Train Epoch: 58 [25600/60000 (43%)]\tLoss: -1928.445508\n",
            "Train Epoch: 58 [26880/60000 (45%)]\tLoss: -1941.805859\n",
            "Train Epoch: 58 [28160/60000 (47%)]\tLoss: -2025.413672\n",
            "Train Epoch: 58 [29440/60000 (49%)]\tLoss: -1945.080664\n",
            "Train Epoch: 58 [30720/60000 (51%)]\tLoss: -1960.036328\n",
            "Train Epoch: 58 [32000/60000 (53%)]\tLoss: -1990.995117\n",
            "Train Epoch: 58 [33280/60000 (55%)]\tLoss: -1956.518164\n",
            "Train Epoch: 58 [34560/60000 (58%)]\tLoss: -2079.850977\n",
            "Train Epoch: 58 [35840/60000 (60%)]\tLoss: -2010.782813\n",
            "Train Epoch: 58 [37120/60000 (62%)]\tLoss: -2055.546875\n",
            "Train Epoch: 58 [38400/60000 (64%)]\tLoss: -1978.710352\n",
            "Train Epoch: 58 [39680/60000 (66%)]\tLoss: -1981.133398\n",
            "Train Epoch: 58 [40960/60000 (68%)]\tLoss: -2085.657617\n",
            "Train Epoch: 58 [42240/60000 (70%)]\tLoss: -2024.424609\n",
            "Train Epoch: 58 [43520/60000 (72%)]\tLoss: -1941.310547\n",
            "Train Epoch: 58 [44800/60000 (75%)]\tLoss: -2016.160938\n",
            "Train Epoch: 58 [46080/60000 (77%)]\tLoss: -2053.934570\n",
            "Train Epoch: 58 [47360/60000 (79%)]\tLoss: -2041.881250\n",
            "Train Epoch: 58 [48640/60000 (81%)]\tLoss: -2009.190039\n",
            "Train Epoch: 58 [49920/60000 (83%)]\tLoss: -2037.540430\n",
            "Train Epoch: 58 [51200/60000 (85%)]\tLoss: -2155.675195\n",
            "Train Epoch: 58 [52480/60000 (87%)]\tLoss: -1999.778516\n",
            "Train Epoch: 58 [53760/60000 (90%)]\tLoss: -1970.804297\n",
            "Train Epoch: 58 [55040/60000 (92%)]\tLoss: -1957.805273\n",
            "Train Epoch: 58 [56320/60000 (94%)]\tLoss: -1942.276953\n",
            "Train Epoch: 58 [57600/60000 (96%)]\tLoss: -1971.261133\n",
            "Train Epoch: 58 [58880/60000 (98%)]\tLoss: -1935.750781\n",
            "====> Epoch: 58 Average loss: -19997.5578\n",
            "Train Epoch: 59 [0/60000 (0%)]\tLoss: -19601.070312\n",
            "Train Epoch: 59 [1280/60000 (2%)]\tLoss: -2231.810352\n",
            "Train Epoch: 59 [2560/60000 (4%)]\tLoss: -1939.670703\n",
            "Train Epoch: 59 [3840/60000 (6%)]\tLoss: -1960.487305\n",
            "Train Epoch: 59 [5120/60000 (9%)]\tLoss: -1990.798438\n",
            "Train Epoch: 59 [6400/60000 (11%)]\tLoss: -1980.140625\n",
            "Train Epoch: 59 [7680/60000 (13%)]\tLoss: -1990.741602\n",
            "Train Epoch: 59 [8960/60000 (15%)]\tLoss: -2046.021094\n",
            "Train Epoch: 59 [10240/60000 (17%)]\tLoss: -2238.768359\n",
            "Train Epoch: 59 [11520/60000 (19%)]\tLoss: -1945.154102\n",
            "Train Epoch: 59 [12800/60000 (21%)]\tLoss: -2026.207617\n",
            "Train Epoch: 59 [14080/60000 (23%)]\tLoss: -2176.097852\n",
            "Train Epoch: 59 [15360/60000 (26%)]\tLoss: -1971.656641\n",
            "Train Epoch: 59 [16640/60000 (28%)]\tLoss: -1948.514453\n",
            "Train Epoch: 59 [17920/60000 (30%)]\tLoss: -1976.395313\n",
            "Train Epoch: 59 [19200/60000 (32%)]\tLoss: -2050.852930\n",
            "Train Epoch: 59 [20480/60000 (34%)]\tLoss: -1923.042773\n",
            "Train Epoch: 59 [21760/60000 (36%)]\tLoss: -2180.516797\n",
            "Train Epoch: 59 [23040/60000 (38%)]\tLoss: -2126.751758\n",
            "Train Epoch: 59 [24320/60000 (41%)]\tLoss: -2035.038867\n",
            "Train Epoch: 59 [25600/60000 (43%)]\tLoss: -1926.195312\n",
            "Train Epoch: 59 [26880/60000 (45%)]\tLoss: -1942.168164\n",
            "Train Epoch: 59 [28160/60000 (47%)]\tLoss: -2005.066016\n",
            "Train Epoch: 59 [29440/60000 (49%)]\tLoss: -1950.240820\n",
            "Train Epoch: 59 [30720/60000 (51%)]\tLoss: -1962.891406\n",
            "Train Epoch: 59 [32000/60000 (53%)]\tLoss: -1971.830469\n",
            "Train Epoch: 59 [33280/60000 (55%)]\tLoss: -1948.218359\n",
            "Train Epoch: 59 [34560/60000 (58%)]\tLoss: -2064.895117\n",
            "Train Epoch: 59 [35840/60000 (60%)]\tLoss: -1994.426367\n",
            "Train Epoch: 59 [37120/60000 (62%)]\tLoss: -2055.982422\n",
            "Train Epoch: 59 [38400/60000 (64%)]\tLoss: -1977.615039\n",
            "Train Epoch: 59 [39680/60000 (66%)]\tLoss: -1979.717187\n",
            "Train Epoch: 59 [40960/60000 (68%)]\tLoss: -2061.351758\n",
            "Train Epoch: 59 [42240/60000 (70%)]\tLoss: -1993.357617\n",
            "Train Epoch: 59 [43520/60000 (72%)]\tLoss: -1928.947656\n",
            "Train Epoch: 59 [44800/60000 (75%)]\tLoss: -1996.385156\n",
            "Train Epoch: 59 [46080/60000 (77%)]\tLoss: -2056.592773\n",
            "Train Epoch: 59 [47360/60000 (79%)]\tLoss: -2043.138672\n",
            "Train Epoch: 59 [48640/60000 (81%)]\tLoss: -2018.828125\n",
            "Train Epoch: 59 [49920/60000 (83%)]\tLoss: -1996.170312\n",
            "Train Epoch: 59 [51200/60000 (85%)]\tLoss: -2156.650000\n",
            "Train Epoch: 59 [52480/60000 (87%)]\tLoss: -1971.708203\n",
            "Train Epoch: 59 [53760/60000 (90%)]\tLoss: -1972.696875\n",
            "Train Epoch: 59 [55040/60000 (92%)]\tLoss: -1943.774805\n",
            "Train Epoch: 59 [56320/60000 (94%)]\tLoss: -1949.598047\n",
            "Train Epoch: 59 [57600/60000 (96%)]\tLoss: -1971.400586\n",
            "Train Epoch: 59 [58880/60000 (98%)]\tLoss: -1911.534766\n",
            "====> Epoch: 59 Average loss: -19999.0892\n",
            "Train Epoch: 60 [0/60000 (0%)]\tLoss: -19615.968750\n",
            "Train Epoch: 60 [1280/60000 (2%)]\tLoss: -2259.350781\n",
            "Train Epoch: 60 [2560/60000 (4%)]\tLoss: -1935.076758\n",
            "Train Epoch: 60 [3840/60000 (6%)]\tLoss: -1958.961328\n",
            "Train Epoch: 60 [5120/60000 (9%)]\tLoss: -2006.082422\n",
            "Train Epoch: 60 [6400/60000 (11%)]\tLoss: -1988.355469\n",
            "Train Epoch: 60 [7680/60000 (13%)]\tLoss: -1991.154883\n",
            "Train Epoch: 60 [8960/60000 (15%)]\tLoss: -2066.132812\n",
            "Train Epoch: 60 [10240/60000 (17%)]\tLoss: -2254.866797\n",
            "Train Epoch: 60 [11520/60000 (19%)]\tLoss: -1941.082422\n",
            "Train Epoch: 60 [12800/60000 (21%)]\tLoss: -1988.867383\n",
            "Train Epoch: 60 [14080/60000 (23%)]\tLoss: -2160.539453\n",
            "Train Epoch: 60 [15360/60000 (26%)]\tLoss: -1969.453516\n",
            "Train Epoch: 60 [16640/60000 (28%)]\tLoss: -1947.072070\n",
            "Train Epoch: 60 [17920/60000 (30%)]\tLoss: -1955.050586\n",
            "Train Epoch: 60 [19200/60000 (32%)]\tLoss: -2037.740039\n",
            "Train Epoch: 60 [20480/60000 (34%)]\tLoss: -1920.183203\n",
            "Train Epoch: 60 [21760/60000 (36%)]\tLoss: -2189.267578\n",
            "Train Epoch: 60 [23040/60000 (38%)]\tLoss: -2123.688477\n",
            "Train Epoch: 60 [24320/60000 (41%)]\tLoss: -2036.885156\n",
            "Train Epoch: 60 [25600/60000 (43%)]\tLoss: -1943.675195\n",
            "Train Epoch: 60 [26880/60000 (45%)]\tLoss: -1940.792187\n",
            "Train Epoch: 60 [28160/60000 (47%)]\tLoss: -2005.727148\n",
            "Train Epoch: 60 [29440/60000 (49%)]\tLoss: -1961.849414\n",
            "Train Epoch: 60 [30720/60000 (51%)]\tLoss: -1970.468945\n",
            "Train Epoch: 60 [32000/60000 (53%)]\tLoss: -1988.522266\n",
            "Train Epoch: 60 [33280/60000 (55%)]\tLoss: -1969.359375\n",
            "Train Epoch: 60 [34560/60000 (58%)]\tLoss: -2005.621289\n",
            "Train Epoch: 60 [35840/60000 (60%)]\tLoss: -1995.922070\n",
            "Train Epoch: 60 [37120/60000 (62%)]\tLoss: -2067.373047\n",
            "Train Epoch: 60 [38400/60000 (64%)]\tLoss: -1968.202930\n",
            "Train Epoch: 60 [39680/60000 (66%)]\tLoss: -1970.907031\n",
            "Train Epoch: 60 [40960/60000 (68%)]\tLoss: -2058.562109\n",
            "Train Epoch: 60 [42240/60000 (70%)]\tLoss: -2005.356445\n",
            "Train Epoch: 60 [43520/60000 (72%)]\tLoss: -1940.300195\n",
            "Train Epoch: 60 [44800/60000 (75%)]\tLoss: -1994.685352\n",
            "Train Epoch: 60 [46080/60000 (77%)]\tLoss: -2045.870117\n",
            "Train Epoch: 60 [47360/60000 (79%)]\tLoss: -2052.253125\n",
            "Train Epoch: 60 [48640/60000 (81%)]\tLoss: -2008.031641\n",
            "Train Epoch: 60 [49920/60000 (83%)]\tLoss: -2022.111133\n",
            "Train Epoch: 60 [51200/60000 (85%)]\tLoss: -2148.391016\n",
            "Train Epoch: 60 [52480/60000 (87%)]\tLoss: -1968.436133\n",
            "Train Epoch: 60 [53760/60000 (90%)]\tLoss: -1969.708398\n",
            "Train Epoch: 60 [55040/60000 (92%)]\tLoss: -1928.073828\n",
            "Train Epoch: 60 [56320/60000 (94%)]\tLoss: -1940.160156\n",
            "Train Epoch: 60 [57600/60000 (96%)]\tLoss: -1977.231445\n",
            "Train Epoch: 60 [58880/60000 (98%)]\tLoss: -1914.251563\n",
            "====> Epoch: 60 Average loss: -19983.6029\n",
            "Train Epoch: 61 [0/60000 (0%)]\tLoss: -19657.875000\n",
            "Train Epoch: 61 [1280/60000 (2%)]\tLoss: -2211.121484\n",
            "Train Epoch: 61 [2560/60000 (4%)]\tLoss: -1935.776953\n",
            "Train Epoch: 61 [3840/60000 (6%)]\tLoss: -1953.573633\n",
            "Train Epoch: 61 [5120/60000 (9%)]\tLoss: -2012.545508\n",
            "Train Epoch: 61 [6400/60000 (11%)]\tLoss: -1972.608984\n",
            "Train Epoch: 61 [7680/60000 (13%)]\tLoss: -1974.791602\n",
            "Train Epoch: 61 [8960/60000 (15%)]\tLoss: -2069.152930\n",
            "Train Epoch: 61 [10240/60000 (17%)]\tLoss: -2257.936914\n",
            "Train Epoch: 61 [11520/60000 (19%)]\tLoss: -1941.059961\n",
            "Train Epoch: 61 [12800/60000 (21%)]\tLoss: -1985.113086\n",
            "Train Epoch: 61 [14080/60000 (23%)]\tLoss: -2178.076172\n",
            "Train Epoch: 61 [15360/60000 (26%)]\tLoss: -1983.750000\n",
            "Train Epoch: 61 [16640/60000 (28%)]\tLoss: -1946.453516\n",
            "Train Epoch: 61 [17920/60000 (30%)]\tLoss: -1946.418359\n",
            "Train Epoch: 61 [19200/60000 (32%)]\tLoss: -2034.583984\n",
            "Train Epoch: 61 [20480/60000 (34%)]\tLoss: -1926.373633\n",
            "Train Epoch: 61 [21760/60000 (36%)]\tLoss: -2159.045117\n",
            "Train Epoch: 61 [23040/60000 (38%)]\tLoss: -2141.411133\n",
            "Train Epoch: 61 [24320/60000 (41%)]\tLoss: -2050.309570\n",
            "Train Epoch: 61 [25600/60000 (43%)]\tLoss: -1923.714258\n",
            "Train Epoch: 61 [26880/60000 (45%)]\tLoss: -1941.873828\n",
            "Train Epoch: 61 [28160/60000 (47%)]\tLoss: -2006.236914\n",
            "Train Epoch: 61 [29440/60000 (49%)]\tLoss: -1942.477930\n",
            "Train Epoch: 61 [30720/60000 (51%)]\tLoss: -1978.083008\n",
            "Train Epoch: 61 [32000/60000 (53%)]\tLoss: -1990.872461\n",
            "Train Epoch: 61 [33280/60000 (55%)]\tLoss: -1965.657813\n",
            "Train Epoch: 61 [34560/60000 (58%)]\tLoss: -2011.319336\n",
            "Train Epoch: 61 [35840/60000 (60%)]\tLoss: -1992.172266\n",
            "Train Epoch: 61 [37120/60000 (62%)]\tLoss: -2052.894922\n",
            "Train Epoch: 61 [38400/60000 (64%)]\tLoss: -1986.088672\n",
            "Train Epoch: 61 [39680/60000 (66%)]\tLoss: -1973.088867\n",
            "Train Epoch: 61 [40960/60000 (68%)]\tLoss: -2045.529297\n",
            "Train Epoch: 61 [42240/60000 (70%)]\tLoss: -2005.894922\n",
            "Train Epoch: 61 [43520/60000 (72%)]\tLoss: -1962.270703\n",
            "Train Epoch: 61 [44800/60000 (75%)]\tLoss: -2016.239648\n",
            "Train Epoch: 61 [46080/60000 (77%)]\tLoss: -2022.119141\n",
            "Train Epoch: 61 [47360/60000 (79%)]\tLoss: -2053.941016\n",
            "Train Epoch: 61 [48640/60000 (81%)]\tLoss: -2029.010156\n",
            "Train Epoch: 61 [49920/60000 (83%)]\tLoss: -1992.109766\n",
            "Train Epoch: 61 [51200/60000 (85%)]\tLoss: -2108.120508\n",
            "Train Epoch: 61 [52480/60000 (87%)]\tLoss: -1967.507812\n",
            "Train Epoch: 61 [53760/60000 (90%)]\tLoss: -1971.910938\n",
            "Train Epoch: 61 [55040/60000 (92%)]\tLoss: -1928.400586\n",
            "Train Epoch: 61 [56320/60000 (94%)]\tLoss: -1949.871484\n",
            "Train Epoch: 61 [57600/60000 (96%)]\tLoss: -1971.130078\n",
            "Train Epoch: 61 [58880/60000 (98%)]\tLoss: -1915.514453\n",
            "====> Epoch: 61 Average loss: -19972.0352\n",
            "Train Epoch: 62 [0/60000 (0%)]\tLoss: -19635.593750\n",
            "Train Epoch: 62 [1280/60000 (2%)]\tLoss: -2211.998047\n",
            "Train Epoch: 62 [2560/60000 (4%)]\tLoss: -1937.570508\n",
            "Train Epoch: 62 [3840/60000 (6%)]\tLoss: -1977.954102\n",
            "Train Epoch: 62 [5120/60000 (9%)]\tLoss: -2005.144141\n",
            "Train Epoch: 62 [6400/60000 (11%)]\tLoss: -1979.996484\n",
            "Train Epoch: 62 [7680/60000 (13%)]\tLoss: -1998.658203\n",
            "Train Epoch: 62 [8960/60000 (15%)]\tLoss: -2062.144141\n",
            "Train Epoch: 62 [10240/60000 (17%)]\tLoss: -2228.768164\n",
            "Train Epoch: 62 [11520/60000 (19%)]\tLoss: -1943.759766\n",
            "Train Epoch: 62 [12800/60000 (21%)]\tLoss: -1990.357812\n",
            "Train Epoch: 62 [14080/60000 (23%)]\tLoss: -2183.929297\n",
            "Train Epoch: 62 [15360/60000 (26%)]\tLoss: -1970.962500\n",
            "Train Epoch: 62 [16640/60000 (28%)]\tLoss: -1946.733008\n",
            "Train Epoch: 62 [17920/60000 (30%)]\tLoss: -1943.738867\n",
            "Train Epoch: 62 [19200/60000 (32%)]\tLoss: -2032.232617\n",
            "Train Epoch: 62 [20480/60000 (34%)]\tLoss: -1923.374414\n",
            "Train Epoch: 62 [21760/60000 (36%)]\tLoss: -2202.267383\n",
            "Train Epoch: 62 [23040/60000 (38%)]\tLoss: -2099.146289\n",
            "Train Epoch: 62 [24320/60000 (41%)]\tLoss: -2044.621680\n",
            "Train Epoch: 62 [25600/60000 (43%)]\tLoss: -1921.946094\n",
            "Train Epoch: 62 [26880/60000 (45%)]\tLoss: -1937.391406\n",
            "Train Epoch: 62 [28160/60000 (47%)]\tLoss: -2021.176172\n",
            "Train Epoch: 62 [29440/60000 (49%)]\tLoss: -1943.722852\n",
            "Train Epoch: 62 [30720/60000 (51%)]\tLoss: -1958.043555\n",
            "Train Epoch: 62 [32000/60000 (53%)]\tLoss: -2012.489844\n",
            "Train Epoch: 62 [33280/60000 (55%)]\tLoss: -1967.541016\n",
            "Train Epoch: 62 [34560/60000 (58%)]\tLoss: -2037.316797\n",
            "Train Epoch: 62 [35840/60000 (60%)]\tLoss: -1992.882422\n",
            "Train Epoch: 62 [37120/60000 (62%)]\tLoss: -2086.006445\n",
            "Train Epoch: 62 [38400/60000 (64%)]\tLoss: -1968.764453\n",
            "Train Epoch: 62 [39680/60000 (66%)]\tLoss: -1973.891602\n",
            "Train Epoch: 62 [40960/60000 (68%)]\tLoss: -2047.590234\n",
            "Train Epoch: 62 [42240/60000 (70%)]\tLoss: -1981.462305\n",
            "Train Epoch: 62 [43520/60000 (72%)]\tLoss: -1944.366797\n",
            "Train Epoch: 62 [44800/60000 (75%)]\tLoss: -1989.524023\n",
            "Train Epoch: 62 [46080/60000 (77%)]\tLoss: -2019.218359\n",
            "Train Epoch: 62 [47360/60000 (79%)]\tLoss: -2038.053711\n",
            "Train Epoch: 62 [48640/60000 (81%)]\tLoss: -2011.299609\n",
            "Train Epoch: 62 [49920/60000 (83%)]\tLoss: -2027.177344\n",
            "Train Epoch: 62 [51200/60000 (85%)]\tLoss: -2121.520703\n",
            "Train Epoch: 62 [52480/60000 (87%)]\tLoss: -1968.403711\n",
            "Train Epoch: 62 [53760/60000 (90%)]\tLoss: -1968.935547\n",
            "Train Epoch: 62 [55040/60000 (92%)]\tLoss: -1947.638477\n",
            "Train Epoch: 62 [56320/60000 (94%)]\tLoss: -1940.084961\n",
            "Train Epoch: 62 [57600/60000 (96%)]\tLoss: -1970.543945\n",
            "Train Epoch: 62 [58880/60000 (98%)]\tLoss: -1917.587500\n",
            "====> Epoch: 62 Average loss: -19985.1862\n",
            "Train Epoch: 63 [0/60000 (0%)]\tLoss: -19719.623047\n",
            "Train Epoch: 63 [1280/60000 (2%)]\tLoss: -2205.181250\n",
            "Train Epoch: 63 [2560/60000 (4%)]\tLoss: -1933.526758\n",
            "Train Epoch: 63 [3840/60000 (6%)]\tLoss: -1961.283398\n",
            "Train Epoch: 63 [5120/60000 (9%)]\tLoss: -1996.518164\n",
            "Train Epoch: 63 [6400/60000 (11%)]\tLoss: -1969.530469\n",
            "Train Epoch: 63 [7680/60000 (13%)]\tLoss: -1973.172852\n",
            "Train Epoch: 63 [8960/60000 (15%)]\tLoss: -2009.550195\n",
            "Train Epoch: 63 [10240/60000 (17%)]\tLoss: -2243.182617\n",
            "Train Epoch: 63 [11520/60000 (19%)]\tLoss: -1963.075391\n",
            "Train Epoch: 63 [12800/60000 (21%)]\tLoss: -1981.714844\n",
            "Train Epoch: 63 [14080/60000 (23%)]\tLoss: -2158.625977\n",
            "Train Epoch: 63 [15360/60000 (26%)]\tLoss: -1970.847070\n",
            "Train Epoch: 63 [16640/60000 (28%)]\tLoss: -1943.311328\n",
            "Train Epoch: 63 [17920/60000 (30%)]\tLoss: -1953.451562\n",
            "Train Epoch: 63 [19200/60000 (32%)]\tLoss: -2049.384570\n",
            "Train Epoch: 63 [20480/60000 (34%)]\tLoss: -1919.775781\n",
            "Train Epoch: 63 [21760/60000 (36%)]\tLoss: -2180.620703\n",
            "Train Epoch: 63 [23040/60000 (38%)]\tLoss: -2091.546875\n",
            "Train Epoch: 63 [24320/60000 (41%)]\tLoss: -2033.161328\n",
            "Train Epoch: 63 [25600/60000 (43%)]\tLoss: -1932.201758\n",
            "Train Epoch: 63 [26880/60000 (45%)]\tLoss: -1942.828125\n",
            "Train Epoch: 63 [28160/60000 (47%)]\tLoss: -2007.019531\n",
            "Train Epoch: 63 [29440/60000 (49%)]\tLoss: -1957.883984\n",
            "Train Epoch: 63 [30720/60000 (51%)]\tLoss: -1966.932227\n",
            "Train Epoch: 63 [32000/60000 (53%)]\tLoss: -1982.544141\n",
            "Train Epoch: 63 [33280/60000 (55%)]\tLoss: -1957.742969\n",
            "Train Epoch: 63 [34560/60000 (58%)]\tLoss: -2033.959375\n",
            "Train Epoch: 63 [35840/60000 (60%)]\tLoss: -1991.089062\n",
            "Train Epoch: 63 [37120/60000 (62%)]\tLoss: -2069.323242\n",
            "Train Epoch: 63 [38400/60000 (64%)]\tLoss: -2002.953320\n",
            "Train Epoch: 63 [39680/60000 (66%)]\tLoss: -1972.596484\n",
            "Train Epoch: 63 [40960/60000 (68%)]\tLoss: -2023.630469\n",
            "Train Epoch: 63 [42240/60000 (70%)]\tLoss: -2004.095898\n",
            "Train Epoch: 63 [43520/60000 (72%)]\tLoss: -1958.535938\n",
            "Train Epoch: 63 [44800/60000 (75%)]\tLoss: -1986.703125\n",
            "Train Epoch: 63 [46080/60000 (77%)]\tLoss: -1997.707227\n",
            "Train Epoch: 63 [47360/60000 (79%)]\tLoss: -2045.332617\n",
            "Train Epoch: 63 [48640/60000 (81%)]\tLoss: -2027.625977\n",
            "Train Epoch: 63 [49920/60000 (83%)]\tLoss: -2014.567578\n",
            "Train Epoch: 63 [51200/60000 (85%)]\tLoss: -2117.324805\n",
            "Train Epoch: 63 [52480/60000 (87%)]\tLoss: -1951.371289\n",
            "Train Epoch: 63 [53760/60000 (90%)]\tLoss: -1967.795312\n",
            "Train Epoch: 63 [55040/60000 (92%)]\tLoss: -1953.099805\n",
            "Train Epoch: 63 [56320/60000 (94%)]\tLoss: -1945.720508\n",
            "Train Epoch: 63 [57600/60000 (96%)]\tLoss: -1973.776367\n",
            "Train Epoch: 63 [58880/60000 (98%)]\tLoss: -1911.825391\n",
            "====> Epoch: 63 Average loss: -19973.2104\n",
            "Train Epoch: 64 [0/60000 (0%)]\tLoss: -19591.906250\n",
            "Train Epoch: 64 [1280/60000 (2%)]\tLoss: -2237.684961\n",
            "Train Epoch: 64 [2560/60000 (4%)]\tLoss: -1933.050977\n",
            "Train Epoch: 64 [3840/60000 (6%)]\tLoss: -1955.160352\n",
            "Train Epoch: 64 [5120/60000 (9%)]\tLoss: -2019.747070\n",
            "Train Epoch: 64 [6400/60000 (11%)]\tLoss: -1945.462305\n",
            "Train Epoch: 64 [7680/60000 (13%)]\tLoss: -1965.094141\n",
            "Train Epoch: 64 [8960/60000 (15%)]\tLoss: -2043.645117\n",
            "Train Epoch: 64 [10240/60000 (17%)]\tLoss: -2253.288867\n",
            "Train Epoch: 64 [11520/60000 (19%)]\tLoss: -1943.357227\n",
            "Train Epoch: 64 [12800/60000 (21%)]\tLoss: -2015.490039\n",
            "Train Epoch: 64 [14080/60000 (23%)]\tLoss: -2160.622656\n",
            "Train Epoch: 64 [15360/60000 (26%)]\tLoss: -1968.846289\n",
            "Train Epoch: 64 [16640/60000 (28%)]\tLoss: -1945.890625\n",
            "Train Epoch: 64 [17920/60000 (30%)]\tLoss: -1953.490820\n",
            "Train Epoch: 64 [19200/60000 (32%)]\tLoss: -2046.786133\n",
            "Train Epoch: 64 [20480/60000 (34%)]\tLoss: -1923.442773\n",
            "Train Epoch: 64 [21760/60000 (36%)]\tLoss: -2211.395312\n",
            "Train Epoch: 64 [23040/60000 (38%)]\tLoss: -2067.774414\n",
            "Train Epoch: 64 [24320/60000 (41%)]\tLoss: -2034.227539\n",
            "Train Epoch: 64 [25600/60000 (43%)]\tLoss: -1944.014648\n",
            "Train Epoch: 64 [26880/60000 (45%)]\tLoss: -1944.367578\n",
            "Train Epoch: 64 [28160/60000 (47%)]\tLoss: -2031.278711\n",
            "Train Epoch: 64 [29440/60000 (49%)]\tLoss: -1974.838477\n",
            "Train Epoch: 64 [30720/60000 (51%)]\tLoss: -1970.997656\n",
            "Train Epoch: 64 [32000/60000 (53%)]\tLoss: -1986.989063\n",
            "Train Epoch: 64 [33280/60000 (55%)]\tLoss: -1960.397070\n",
            "Train Epoch: 64 [34560/60000 (58%)]\tLoss: -2037.281445\n",
            "Train Epoch: 64 [35840/60000 (60%)]\tLoss: -1990.947266\n",
            "Train Epoch: 64 [37120/60000 (62%)]\tLoss: -2067.756055\n",
            "Train Epoch: 64 [38400/60000 (64%)]\tLoss: -1969.446875\n",
            "Train Epoch: 64 [39680/60000 (66%)]\tLoss: -1964.772656\n",
            "Train Epoch: 64 [40960/60000 (68%)]\tLoss: -2035.926953\n",
            "Train Epoch: 64 [42240/60000 (70%)]\tLoss: -2011.847266\n",
            "Train Epoch: 64 [43520/60000 (72%)]\tLoss: -1956.893164\n",
            "Train Epoch: 64 [44800/60000 (75%)]\tLoss: -2015.292383\n",
            "Train Epoch: 64 [46080/60000 (77%)]\tLoss: -2002.259375\n",
            "Train Epoch: 64 [47360/60000 (79%)]\tLoss: -2033.977734\n",
            "Train Epoch: 64 [48640/60000 (81%)]\tLoss: -2028.339844\n",
            "Train Epoch: 64 [49920/60000 (83%)]\tLoss: -2011.032617\n",
            "Train Epoch: 64 [51200/60000 (85%)]\tLoss: -2094.302539\n",
            "Train Epoch: 64 [52480/60000 (87%)]\tLoss: -1944.827344\n",
            "Train Epoch: 64 [53760/60000 (90%)]\tLoss: -1968.531445\n",
            "Train Epoch: 64 [55040/60000 (92%)]\tLoss: -1937.407813\n",
            "Train Epoch: 64 [56320/60000 (94%)]\tLoss: -1946.496484\n",
            "Train Epoch: 64 [57600/60000 (96%)]\tLoss: -1965.487500\n",
            "Train Epoch: 64 [58880/60000 (98%)]\tLoss: -1915.555273\n",
            "====> Epoch: 64 Average loss: -19964.5345\n",
            "Train Epoch: 65 [0/60000 (0%)]\tLoss: -19581.328125\n",
            "Train Epoch: 65 [1280/60000 (2%)]\tLoss: -2204.958398\n",
            "Train Epoch: 65 [2560/60000 (4%)]\tLoss: -1949.975586\n",
            "Train Epoch: 65 [3840/60000 (6%)]\tLoss: -1979.868164\n",
            "Train Epoch: 65 [5120/60000 (9%)]\tLoss: -2010.466016\n",
            "Train Epoch: 65 [6400/60000 (11%)]\tLoss: -1972.651367\n",
            "Train Epoch: 65 [7680/60000 (13%)]\tLoss: -1967.978516\n",
            "Train Epoch: 65 [8960/60000 (15%)]\tLoss: -2031.880078\n",
            "Train Epoch: 65 [10240/60000 (17%)]\tLoss: -2239.623047\n",
            "Train Epoch: 65 [11520/60000 (19%)]\tLoss: -1942.520508\n",
            "Train Epoch: 65 [12800/60000 (21%)]\tLoss: -2002.598242\n",
            "Train Epoch: 65 [14080/60000 (23%)]\tLoss: -2179.934961\n",
            "Train Epoch: 65 [15360/60000 (26%)]\tLoss: -1961.302930\n",
            "Train Epoch: 65 [16640/60000 (28%)]\tLoss: -1945.260547\n",
            "Train Epoch: 65 [17920/60000 (30%)]\tLoss: -1950.709766\n",
            "Train Epoch: 65 [19200/60000 (32%)]\tLoss: -2031.553516\n",
            "Train Epoch: 65 [20480/60000 (34%)]\tLoss: -1926.101953\n",
            "Train Epoch: 65 [21760/60000 (36%)]\tLoss: -2227.075195\n",
            "Train Epoch: 65 [23040/60000 (38%)]\tLoss: -2094.924805\n",
            "Train Epoch: 65 [24320/60000 (41%)]\tLoss: -2033.440430\n",
            "Train Epoch: 65 [25600/60000 (43%)]\tLoss: -1929.913086\n",
            "Train Epoch: 65 [26880/60000 (45%)]\tLoss: -1944.740625\n",
            "Train Epoch: 65 [28160/60000 (47%)]\tLoss: -2019.932031\n",
            "Train Epoch: 65 [29440/60000 (49%)]\tLoss: -1958.343750\n",
            "Train Epoch: 65 [30720/60000 (51%)]\tLoss: -1983.989648\n",
            "Train Epoch: 65 [32000/60000 (53%)]\tLoss: -1983.723438\n",
            "Train Epoch: 65 [33280/60000 (55%)]\tLoss: -1945.619141\n",
            "Train Epoch: 65 [34560/60000 (58%)]\tLoss: -2000.109180\n",
            "Train Epoch: 65 [35840/60000 (60%)]\tLoss: -1992.588281\n",
            "Train Epoch: 65 [37120/60000 (62%)]\tLoss: -2079.782422\n",
            "Train Epoch: 65 [38400/60000 (64%)]\tLoss: -1953.683203\n",
            "Train Epoch: 65 [39680/60000 (66%)]\tLoss: -1968.229297\n",
            "Train Epoch: 65 [40960/60000 (68%)]\tLoss: -2036.809375\n",
            "Train Epoch: 65 [42240/60000 (70%)]\tLoss: -1990.280469\n",
            "Train Epoch: 65 [43520/60000 (72%)]\tLoss: -1963.604492\n",
            "Train Epoch: 65 [44800/60000 (75%)]\tLoss: -2011.000391\n",
            "Train Epoch: 65 [46080/60000 (77%)]\tLoss: -2013.511328\n",
            "Train Epoch: 65 [47360/60000 (79%)]\tLoss: -2065.643164\n",
            "Train Epoch: 65 [48640/60000 (81%)]\tLoss: -2015.185156\n",
            "Train Epoch: 65 [49920/60000 (83%)]\tLoss: -2005.104883\n",
            "Train Epoch: 65 [51200/60000 (85%)]\tLoss: -2125.463086\n",
            "Train Epoch: 65 [52480/60000 (87%)]\tLoss: -1937.551367\n",
            "Train Epoch: 65 [53760/60000 (90%)]\tLoss: -1970.845898\n",
            "Train Epoch: 65 [55040/60000 (92%)]\tLoss: -1949.141602\n",
            "Train Epoch: 65 [56320/60000 (94%)]\tLoss: -1942.934961\n",
            "Train Epoch: 65 [57600/60000 (96%)]\tLoss: -1975.017188\n",
            "Train Epoch: 65 [58880/60000 (98%)]\tLoss: -1913.265625\n",
            "====> Epoch: 65 Average loss: -19977.7680\n",
            "Train Epoch: 66 [0/60000 (0%)]\tLoss: -19618.794922\n",
            "Train Epoch: 66 [1280/60000 (2%)]\tLoss: -2201.331055\n",
            "Train Epoch: 66 [2560/60000 (4%)]\tLoss: -1952.960547\n",
            "Train Epoch: 66 [3840/60000 (6%)]\tLoss: -1961.356836\n",
            "Train Epoch: 66 [5120/60000 (9%)]\tLoss: -2057.019531\n",
            "Train Epoch: 66 [6400/60000 (11%)]\tLoss: -1940.996289\n",
            "Train Epoch: 66 [7680/60000 (13%)]\tLoss: -1975.182422\n",
            "Train Epoch: 66 [8960/60000 (15%)]\tLoss: -2059.658984\n",
            "Train Epoch: 66 [10240/60000 (17%)]\tLoss: -2244.039648\n",
            "Train Epoch: 66 [11520/60000 (19%)]\tLoss: -1946.718750\n",
            "Train Epoch: 66 [12800/60000 (21%)]\tLoss: -1994.289062\n",
            "Train Epoch: 66 [14080/60000 (23%)]\tLoss: -2186.874414\n",
            "Train Epoch: 66 [15360/60000 (26%)]\tLoss: -1961.026562\n",
            "Train Epoch: 66 [16640/60000 (28%)]\tLoss: -1942.652734\n",
            "Train Epoch: 66 [17920/60000 (30%)]\tLoss: -1946.804102\n",
            "Train Epoch: 66 [19200/60000 (32%)]\tLoss: -2032.636719\n",
            "Train Epoch: 66 [20480/60000 (34%)]\tLoss: -1928.595898\n",
            "Train Epoch: 66 [21760/60000 (36%)]\tLoss: -2187.235742\n",
            "Train Epoch: 66 [23040/60000 (38%)]\tLoss: -2138.070508\n",
            "Train Epoch: 66 [24320/60000 (41%)]\tLoss: -2038.669336\n",
            "Train Epoch: 66 [25600/60000 (43%)]\tLoss: -1940.450195\n",
            "Train Epoch: 66 [26880/60000 (45%)]\tLoss: -1944.850977\n",
            "Train Epoch: 66 [28160/60000 (47%)]\tLoss: -2046.679102\n",
            "Train Epoch: 66 [29440/60000 (49%)]\tLoss: -1955.624023\n",
            "Train Epoch: 66 [30720/60000 (51%)]\tLoss: -1961.751758\n",
            "Train Epoch: 66 [32000/60000 (53%)]\tLoss: -1990.519141\n",
            "Train Epoch: 66 [33280/60000 (55%)]\tLoss: -1940.640430\n",
            "Train Epoch: 66 [34560/60000 (58%)]\tLoss: -1966.439844\n",
            "Train Epoch: 66 [35840/60000 (60%)]\tLoss: -1990.158203\n",
            "Train Epoch: 66 [37120/60000 (62%)]\tLoss: -2094.587109\n",
            "Train Epoch: 66 [38400/60000 (64%)]\tLoss: -1955.628906\n",
            "Train Epoch: 66 [39680/60000 (66%)]\tLoss: -1969.575586\n",
            "Train Epoch: 66 [40960/60000 (68%)]\tLoss: -2055.553125\n",
            "Train Epoch: 66 [42240/60000 (70%)]\tLoss: -1976.517578\n",
            "Train Epoch: 66 [43520/60000 (72%)]\tLoss: -1975.098828\n",
            "Train Epoch: 66 [44800/60000 (75%)]\tLoss: -2023.359961\n",
            "Train Epoch: 66 [46080/60000 (77%)]\tLoss: -2011.028711\n",
            "Train Epoch: 66 [47360/60000 (79%)]\tLoss: -2075.910352\n",
            "Train Epoch: 66 [48640/60000 (81%)]\tLoss: -2020.793555\n",
            "Train Epoch: 66 [49920/60000 (83%)]\tLoss: -2023.965234\n",
            "Train Epoch: 66 [51200/60000 (85%)]\tLoss: -2110.782422\n",
            "Train Epoch: 66 [52480/60000 (87%)]\tLoss: -1951.816406\n",
            "Train Epoch: 66 [53760/60000 (90%)]\tLoss: -1970.020313\n",
            "Train Epoch: 66 [55040/60000 (92%)]\tLoss: -1940.530469\n",
            "Train Epoch: 66 [56320/60000 (94%)]\tLoss: -1956.726562\n",
            "Train Epoch: 66 [57600/60000 (96%)]\tLoss: -1962.780664\n",
            "Train Epoch: 66 [58880/60000 (98%)]\tLoss: -1926.290039\n",
            "====> Epoch: 66 Average loss: -19973.4343\n",
            "Train Epoch: 67 [0/60000 (0%)]\tLoss: -19645.587891\n",
            "Train Epoch: 67 [1280/60000 (2%)]\tLoss: -2191.190039\n",
            "Train Epoch: 67 [2560/60000 (4%)]\tLoss: -1951.019531\n",
            "Train Epoch: 67 [3840/60000 (6%)]\tLoss: -1971.440430\n",
            "Train Epoch: 67 [5120/60000 (9%)]\tLoss: -2080.262695\n",
            "Train Epoch: 67 [6400/60000 (11%)]\tLoss: -1968.347656\n",
            "Train Epoch: 67 [7680/60000 (13%)]\tLoss: -1990.364844\n",
            "Train Epoch: 67 [8960/60000 (15%)]\tLoss: -2032.473242\n",
            "Train Epoch: 67 [10240/60000 (17%)]\tLoss: -2254.447070\n",
            "Train Epoch: 67 [11520/60000 (19%)]\tLoss: -1949.973633\n",
            "Train Epoch: 67 [12800/60000 (21%)]\tLoss: -1976.865430\n",
            "Train Epoch: 67 [14080/60000 (23%)]\tLoss: -2191.017578\n",
            "Train Epoch: 67 [15360/60000 (26%)]\tLoss: -1966.180859\n",
            "Train Epoch: 67 [16640/60000 (28%)]\tLoss: -1943.495508\n",
            "Train Epoch: 67 [17920/60000 (30%)]\tLoss: -1957.204688\n",
            "Train Epoch: 67 [19200/60000 (32%)]\tLoss: -2054.883203\n",
            "Train Epoch: 67 [20480/60000 (34%)]\tLoss: -1944.562695\n",
            "Train Epoch: 67 [21760/60000 (36%)]\tLoss: -2223.425391\n",
            "Train Epoch: 67 [23040/60000 (38%)]\tLoss: -2118.221680\n",
            "Train Epoch: 67 [24320/60000 (41%)]\tLoss: -2035.665625\n",
            "Train Epoch: 67 [25600/60000 (43%)]\tLoss: -1927.764648\n",
            "Train Epoch: 67 [26880/60000 (45%)]\tLoss: -1955.026172\n",
            "Train Epoch: 67 [28160/60000 (47%)]\tLoss: -2059.525000\n",
            "Train Epoch: 67 [29440/60000 (49%)]\tLoss: -1951.544727\n",
            "Train Epoch: 67 [30720/60000 (51%)]\tLoss: -1977.996289\n",
            "Train Epoch: 67 [32000/60000 (53%)]\tLoss: -2036.088867\n",
            "Train Epoch: 67 [33280/60000 (55%)]\tLoss: -1947.475781\n",
            "Train Epoch: 67 [34560/60000 (58%)]\tLoss: -1996.013672\n",
            "Train Epoch: 67 [35840/60000 (60%)]\tLoss: -1987.854883\n",
            "Train Epoch: 67 [37120/60000 (62%)]\tLoss: -2093.146484\n",
            "Train Epoch: 67 [38400/60000 (64%)]\tLoss: -1963.045898\n",
            "Train Epoch: 67 [39680/60000 (66%)]\tLoss: -1971.652930\n",
            "Train Epoch: 67 [40960/60000 (68%)]\tLoss: -2038.789453\n",
            "Train Epoch: 67 [42240/60000 (70%)]\tLoss: -1993.193750\n",
            "Train Epoch: 67 [43520/60000 (72%)]\tLoss: -1944.117383\n",
            "Train Epoch: 67 [44800/60000 (75%)]\tLoss: -1989.156055\n",
            "Train Epoch: 67 [46080/60000 (77%)]\tLoss: -2017.592578\n",
            "Train Epoch: 67 [47360/60000 (79%)]\tLoss: -2019.369727\n",
            "Train Epoch: 67 [48640/60000 (81%)]\tLoss: -2007.912695\n",
            "Train Epoch: 67 [49920/60000 (83%)]\tLoss: -2015.211328\n",
            "Train Epoch: 67 [51200/60000 (85%)]\tLoss: -2093.614648\n",
            "Train Epoch: 67 [52480/60000 (87%)]\tLoss: -1937.099414\n",
            "Train Epoch: 67 [53760/60000 (90%)]\tLoss: -1965.376953\n",
            "Train Epoch: 67 [55040/60000 (92%)]\tLoss: -1940.859961\n",
            "Train Epoch: 67 [56320/60000 (94%)]\tLoss: -1941.950977\n",
            "Train Epoch: 67 [57600/60000 (96%)]\tLoss: -1966.625977\n",
            "Train Epoch: 67 [58880/60000 (98%)]\tLoss: -1906.309961\n",
            "====> Epoch: 67 Average loss: -19953.6808\n",
            "Train Epoch: 68 [0/60000 (0%)]\tLoss: -19547.361328\n",
            "Train Epoch: 68 [1280/60000 (2%)]\tLoss: -2168.107031\n",
            "Train Epoch: 68 [2560/60000 (4%)]\tLoss: -1930.867383\n",
            "Train Epoch: 68 [3840/60000 (6%)]\tLoss: -1961.712500\n",
            "Train Epoch: 68 [5120/60000 (9%)]\tLoss: -2031.917578\n",
            "Train Epoch: 68 [6400/60000 (11%)]\tLoss: -1952.660352\n",
            "Train Epoch: 68 [7680/60000 (13%)]\tLoss: -1990.191211\n",
            "Train Epoch: 68 [8960/60000 (15%)]\tLoss: -1987.525391\n",
            "Train Epoch: 68 [10240/60000 (17%)]\tLoss: -2242.274414\n",
            "Train Epoch: 68 [11520/60000 (19%)]\tLoss: -1944.681836\n",
            "Train Epoch: 68 [12800/60000 (21%)]\tLoss: -1988.930469\n",
            "Train Epoch: 68 [14080/60000 (23%)]\tLoss: -2158.629297\n",
            "Train Epoch: 68 [15360/60000 (26%)]\tLoss: -1966.669727\n",
            "Train Epoch: 68 [16640/60000 (28%)]\tLoss: -1943.836719\n",
            "Train Epoch: 68 [17920/60000 (30%)]\tLoss: -1959.975586\n",
            "Train Epoch: 68 [19200/60000 (32%)]\tLoss: -2045.709766\n",
            "Train Epoch: 68 [20480/60000 (34%)]\tLoss: -1918.258789\n",
            "Train Epoch: 68 [21760/60000 (36%)]\tLoss: -2185.856055\n",
            "Train Epoch: 68 [23040/60000 (38%)]\tLoss: -2084.073438\n",
            "Train Epoch: 68 [24320/60000 (41%)]\tLoss: -2009.174805\n",
            "Train Epoch: 68 [25600/60000 (43%)]\tLoss: -1935.550195\n",
            "Train Epoch: 68 [26880/60000 (45%)]\tLoss: -1940.599609\n",
            "Train Epoch: 68 [28160/60000 (47%)]\tLoss: -1988.997461\n",
            "Train Epoch: 68 [29440/60000 (49%)]\tLoss: -1948.654883\n",
            "Train Epoch: 68 [30720/60000 (51%)]\tLoss: -1951.513867\n",
            "Train Epoch: 68 [32000/60000 (53%)]\tLoss: -2005.573242\n",
            "Train Epoch: 68 [33280/60000 (55%)]\tLoss: -1949.872266\n",
            "Train Epoch: 68 [34560/60000 (58%)]\tLoss: -2000.828516\n",
            "Train Epoch: 68 [35840/60000 (60%)]\tLoss: -1995.241211\n",
            "Train Epoch: 68 [37120/60000 (62%)]\tLoss: -2073.329102\n",
            "Train Epoch: 68 [38400/60000 (64%)]\tLoss: -1966.341016\n",
            "Train Epoch: 68 [39680/60000 (66%)]\tLoss: -1970.701367\n",
            "Train Epoch: 68 [40960/60000 (68%)]\tLoss: -2017.757227\n",
            "Train Epoch: 68 [42240/60000 (70%)]\tLoss: -1975.392188\n",
            "Train Epoch: 68 [43520/60000 (72%)]\tLoss: -1961.476172\n",
            "Train Epoch: 68 [44800/60000 (75%)]\tLoss: -1977.513281\n",
            "Train Epoch: 68 [46080/60000 (77%)]\tLoss: -1958.840234\n",
            "Train Epoch: 68 [47360/60000 (79%)]\tLoss: -2021.995898\n",
            "Train Epoch: 68 [48640/60000 (81%)]\tLoss: -2023.032031\n",
            "Train Epoch: 68 [49920/60000 (83%)]\tLoss: -2013.602930\n",
            "Train Epoch: 68 [51200/60000 (85%)]\tLoss: -2109.749414\n",
            "Train Epoch: 68 [52480/60000 (87%)]\tLoss: -1940.154297\n",
            "Train Epoch: 68 [53760/60000 (90%)]\tLoss: -1963.775391\n",
            "Train Epoch: 68 [55040/60000 (92%)]\tLoss: -1943.645313\n",
            "Train Epoch: 68 [56320/60000 (94%)]\tLoss: -1941.149805\n",
            "Train Epoch: 68 [57600/60000 (96%)]\tLoss: -1966.149219\n",
            "Train Epoch: 68 [58880/60000 (98%)]\tLoss: -1907.188477\n",
            "====> Epoch: 68 Average loss: -19872.7328\n",
            "Train Epoch: 69 [0/60000 (0%)]\tLoss: -19658.789062\n",
            "Train Epoch: 69 [1280/60000 (2%)]\tLoss: -2195.745508\n",
            "Train Epoch: 69 [2560/60000 (4%)]\tLoss: -1935.159961\n",
            "Train Epoch: 69 [3840/60000 (6%)]\tLoss: -1946.929492\n",
            "Train Epoch: 69 [5120/60000 (9%)]\tLoss: -2035.459961\n",
            "Train Epoch: 69 [6400/60000 (11%)]\tLoss: -1971.048047\n",
            "Train Epoch: 69 [7680/60000 (13%)]\tLoss: -1952.288281\n",
            "Train Epoch: 69 [8960/60000 (15%)]\tLoss: -2003.589844\n",
            "Train Epoch: 69 [10240/60000 (17%)]\tLoss: -2241.670312\n",
            "Train Epoch: 69 [11520/60000 (19%)]\tLoss: -1946.505469\n",
            "Train Epoch: 69 [12800/60000 (21%)]\tLoss: -1997.796680\n",
            "Train Epoch: 69 [14080/60000 (23%)]\tLoss: -2173.505273\n",
            "Train Epoch: 69 [15360/60000 (26%)]\tLoss: -1978.966602\n",
            "Train Epoch: 69 [16640/60000 (28%)]\tLoss: -1942.537891\n",
            "Train Epoch: 69 [17920/60000 (30%)]\tLoss: -1977.767578\n",
            "Train Epoch: 69 [19200/60000 (32%)]\tLoss: -2054.431641\n",
            "Train Epoch: 69 [20480/60000 (34%)]\tLoss: -1926.698242\n",
            "Train Epoch: 69 [21760/60000 (36%)]\tLoss: -2224.650391\n",
            "Train Epoch: 69 [23040/60000 (38%)]\tLoss: -2092.462109\n",
            "Train Epoch: 69 [24320/60000 (41%)]\tLoss: -2029.358984\n",
            "Train Epoch: 69 [25600/60000 (43%)]\tLoss: -1919.209180\n",
            "Train Epoch: 69 [26880/60000 (45%)]\tLoss: -1942.744727\n",
            "Train Epoch: 69 [28160/60000 (47%)]\tLoss: -2006.486914\n",
            "Train Epoch: 69 [29440/60000 (49%)]\tLoss: -1949.582422\n",
            "Train Epoch: 69 [30720/60000 (51%)]\tLoss: -1950.144922\n",
            "Train Epoch: 69 [32000/60000 (53%)]\tLoss: -2001.130469\n",
            "Train Epoch: 69 [33280/60000 (55%)]\tLoss: -1949.463086\n",
            "Train Epoch: 69 [34560/60000 (58%)]\tLoss: -1997.174609\n",
            "Train Epoch: 69 [35840/60000 (60%)]\tLoss: -1991.963672\n",
            "Train Epoch: 69 [37120/60000 (62%)]\tLoss: -2088.859375\n",
            "Train Epoch: 69 [38400/60000 (64%)]\tLoss: -1998.424023\n",
            "Train Epoch: 69 [39680/60000 (66%)]\tLoss: -1972.891602\n",
            "Train Epoch: 69 [40960/60000 (68%)]\tLoss: -2007.987109\n",
            "Train Epoch: 69 [42240/60000 (70%)]\tLoss: -1982.687500\n",
            "Train Epoch: 69 [43520/60000 (72%)]\tLoss: -1948.941797\n",
            "Train Epoch: 69 [44800/60000 (75%)]\tLoss: -1989.583398\n",
            "Train Epoch: 69 [46080/60000 (77%)]\tLoss: -2008.534375\n",
            "Train Epoch: 69 [47360/60000 (79%)]\tLoss: -2002.342773\n",
            "Train Epoch: 69 [48640/60000 (81%)]\tLoss: -2001.731055\n",
            "Train Epoch: 69 [49920/60000 (83%)]\tLoss: -2017.292383\n",
            "Train Epoch: 69 [51200/60000 (85%)]\tLoss: -2134.178516\n",
            "Train Epoch: 69 [52480/60000 (87%)]\tLoss: -1941.108984\n",
            "Train Epoch: 69 [53760/60000 (90%)]\tLoss: -1965.836328\n",
            "Train Epoch: 69 [55040/60000 (92%)]\tLoss: -1938.879688\n",
            "Train Epoch: 69 [56320/60000 (94%)]\tLoss: -1948.733398\n",
            "Train Epoch: 69 [57600/60000 (96%)]\tLoss: -1968.322656\n",
            "Train Epoch: 69 [58880/60000 (98%)]\tLoss: -1902.740039\n",
            "====> Epoch: 69 Average loss: -19909.7870\n",
            "Train Epoch: 70 [0/60000 (0%)]\tLoss: -19621.996094\n",
            "Train Epoch: 70 [1280/60000 (2%)]\tLoss: -2154.869336\n",
            "Train Epoch: 70 [2560/60000 (4%)]\tLoss: -1931.220898\n",
            "Train Epoch: 70 [3840/60000 (6%)]\tLoss: -1977.675977\n",
            "Train Epoch: 70 [5120/60000 (9%)]\tLoss: -2054.549023\n",
            "Train Epoch: 70 [6400/60000 (11%)]\tLoss: -1964.384375\n",
            "Train Epoch: 70 [7680/60000 (13%)]\tLoss: -1966.232812\n",
            "Train Epoch: 70 [8960/60000 (15%)]\tLoss: -2055.546680\n",
            "Train Epoch: 70 [10240/60000 (17%)]\tLoss: -2266.026172\n",
            "Train Epoch: 70 [11520/60000 (19%)]\tLoss: -1939.589258\n",
            "Train Epoch: 70 [12800/60000 (21%)]\tLoss: -1988.103125\n",
            "Train Epoch: 70 [14080/60000 (23%)]\tLoss: -2160.864062\n",
            "Train Epoch: 70 [15360/60000 (26%)]\tLoss: -1963.704297\n",
            "Train Epoch: 70 [16640/60000 (28%)]\tLoss: -1942.842383\n",
            "Train Epoch: 70 [17920/60000 (30%)]\tLoss: -1976.270117\n",
            "Train Epoch: 70 [19200/60000 (32%)]\tLoss: -2103.303320\n",
            "Train Epoch: 70 [20480/60000 (34%)]\tLoss: -1915.417969\n",
            "Train Epoch: 70 [21760/60000 (36%)]\tLoss: -2222.628711\n",
            "Train Epoch: 70 [23040/60000 (38%)]\tLoss: -2078.677539\n",
            "Train Epoch: 70 [24320/60000 (41%)]\tLoss: -2034.192969\n",
            "Train Epoch: 70 [25600/60000 (43%)]\tLoss: -1924.028125\n",
            "Train Epoch: 70 [26880/60000 (45%)]\tLoss: -1947.962695\n",
            "Train Epoch: 70 [28160/60000 (47%)]\tLoss: -1992.548633\n",
            "Train Epoch: 70 [29440/60000 (49%)]\tLoss: -1951.975195\n",
            "Train Epoch: 70 [30720/60000 (51%)]\tLoss: -1965.476367\n",
            "Train Epoch: 70 [32000/60000 (53%)]\tLoss: -2020.368555\n",
            "Train Epoch: 70 [33280/60000 (55%)]\tLoss: -1954.537305\n",
            "Train Epoch: 70 [34560/60000 (58%)]\tLoss: -2046.390625\n",
            "Train Epoch: 70 [35840/60000 (60%)]\tLoss: -1990.290820\n",
            "Train Epoch: 70 [37120/60000 (62%)]\tLoss: -2067.268945\n",
            "Train Epoch: 70 [38400/60000 (64%)]\tLoss: -1979.960352\n",
            "Train Epoch: 70 [39680/60000 (66%)]\tLoss: -1977.219336\n",
            "Train Epoch: 70 [40960/60000 (68%)]\tLoss: -2037.706836\n",
            "Train Epoch: 70 [42240/60000 (70%)]\tLoss: -2008.612305\n",
            "Train Epoch: 70 [43520/60000 (72%)]\tLoss: -1941.932812\n",
            "Train Epoch: 70 [44800/60000 (75%)]\tLoss: -1986.072266\n",
            "Train Epoch: 70 [46080/60000 (77%)]\tLoss: -2002.408008\n",
            "Train Epoch: 70 [47360/60000 (79%)]\tLoss: -2040.864063\n",
            "Train Epoch: 70 [48640/60000 (81%)]\tLoss: -2017.704492\n",
            "Train Epoch: 70 [49920/60000 (83%)]\tLoss: -2016.310156\n",
            "Train Epoch: 70 [51200/60000 (85%)]\tLoss: -2104.512500\n",
            "Train Epoch: 70 [52480/60000 (87%)]\tLoss: -1939.001953\n",
            "Train Epoch: 70 [53760/60000 (90%)]\tLoss: -1970.138867\n",
            "Train Epoch: 70 [55040/60000 (92%)]\tLoss: -1949.096680\n",
            "Train Epoch: 70 [56320/60000 (94%)]\tLoss: -1938.426953\n",
            "Train Epoch: 70 [57600/60000 (96%)]\tLoss: -1975.944727\n",
            "Train Epoch: 70 [58880/60000 (98%)]\tLoss: -1913.127930\n",
            "====> Epoch: 70 Average loss: -19950.5310\n",
            "Train Epoch: 71 [0/60000 (0%)]\tLoss: -19706.025391\n",
            "Train Epoch: 71 [1280/60000 (2%)]\tLoss: -2188.312305\n",
            "Train Epoch: 71 [2560/60000 (4%)]\tLoss: -1932.397461\n",
            "Train Epoch: 71 [3840/60000 (6%)]\tLoss: -1985.550586\n",
            "Train Epoch: 71 [5120/60000 (9%)]\tLoss: -2050.377734\n",
            "Train Epoch: 71 [6400/60000 (11%)]\tLoss: -1985.059570\n",
            "Train Epoch: 71 [7680/60000 (13%)]\tLoss: -1951.099805\n",
            "Train Epoch: 71 [8960/60000 (15%)]\tLoss: -2046.534375\n",
            "Train Epoch: 71 [10240/60000 (17%)]\tLoss: -2256.136914\n",
            "Train Epoch: 71 [11520/60000 (19%)]\tLoss: -1959.858398\n",
            "Train Epoch: 71 [12800/60000 (21%)]\tLoss: -2011.944922\n",
            "Train Epoch: 71 [14080/60000 (23%)]\tLoss: -2121.991602\n",
            "Train Epoch: 71 [15360/60000 (26%)]\tLoss: -1953.699805\n",
            "Train Epoch: 71 [16640/60000 (28%)]\tLoss: -1946.277344\n",
            "Train Epoch: 71 [17920/60000 (30%)]\tLoss: -1969.877344\n",
            "Train Epoch: 71 [19200/60000 (32%)]\tLoss: -2098.942969\n",
            "Train Epoch: 71 [20480/60000 (34%)]\tLoss: -1910.087695\n",
            "Train Epoch: 71 [21760/60000 (36%)]\tLoss: -2252.246094\n",
            "Train Epoch: 71 [23040/60000 (38%)]\tLoss: -2080.058594\n",
            "Train Epoch: 71 [24320/60000 (41%)]\tLoss: -2076.477344\n",
            "Train Epoch: 71 [25600/60000 (43%)]\tLoss: -1929.558789\n",
            "Train Epoch: 71 [26880/60000 (45%)]\tLoss: -1948.632227\n",
            "Train Epoch: 71 [28160/60000 (47%)]\tLoss: -2026.417773\n",
            "Train Epoch: 71 [29440/60000 (49%)]\tLoss: -1941.861523\n",
            "Train Epoch: 71 [30720/60000 (51%)]\tLoss: -1960.641992\n",
            "Train Epoch: 71 [32000/60000 (53%)]\tLoss: -2010.872656\n",
            "Train Epoch: 71 [33280/60000 (55%)]\tLoss: -1976.065430\n",
            "Train Epoch: 71 [34560/60000 (58%)]\tLoss: -2057.328906\n",
            "Train Epoch: 71 [35840/60000 (60%)]\tLoss: -1989.660742\n",
            "Train Epoch: 71 [37120/60000 (62%)]\tLoss: -2064.952734\n",
            "Train Epoch: 71 [38400/60000 (64%)]\tLoss: -1971.696094\n",
            "Train Epoch: 71 [39680/60000 (66%)]\tLoss: -1971.860352\n",
            "Train Epoch: 71 [40960/60000 (68%)]\tLoss: -2052.100391\n",
            "Train Epoch: 71 [42240/60000 (70%)]\tLoss: -1990.398047\n",
            "Train Epoch: 71 [43520/60000 (72%)]\tLoss: -1949.640039\n",
            "Train Epoch: 71 [44800/60000 (75%)]\tLoss: -1997.889648\n",
            "Train Epoch: 71 [46080/60000 (77%)]\tLoss: -1989.900977\n",
            "Train Epoch: 71 [47360/60000 (79%)]\tLoss: -2051.296484\n",
            "Train Epoch: 71 [48640/60000 (81%)]\tLoss: -2029.422266\n",
            "Train Epoch: 71 [49920/60000 (83%)]\tLoss: -2009.076172\n",
            "Train Epoch: 71 [51200/60000 (85%)]\tLoss: -2125.395703\n",
            "Train Epoch: 71 [52480/60000 (87%)]\tLoss: -1952.568750\n",
            "Train Epoch: 71 [53760/60000 (90%)]\tLoss: -1965.668555\n",
            "Train Epoch: 71 [55040/60000 (92%)]\tLoss: -1956.233789\n",
            "Train Epoch: 71 [56320/60000 (94%)]\tLoss: -1939.558984\n",
            "Train Epoch: 71 [57600/60000 (96%)]\tLoss: -1960.677734\n",
            "Train Epoch: 71 [58880/60000 (98%)]\tLoss: -1912.744141\n",
            "====> Epoch: 71 Average loss: -20005.5743\n",
            "Train Epoch: 72 [0/60000 (0%)]\tLoss: -19652.041016\n",
            "Train Epoch: 72 [1280/60000 (2%)]\tLoss: -2212.267773\n",
            "Train Epoch: 72 [2560/60000 (4%)]\tLoss: -1932.395508\n",
            "Train Epoch: 72 [3840/60000 (6%)]\tLoss: -1968.062695\n",
            "Train Epoch: 72 [5120/60000 (9%)]\tLoss: -2045.507422\n",
            "Train Epoch: 72 [6400/60000 (11%)]\tLoss: -1969.483398\n",
            "Train Epoch: 72 [7680/60000 (13%)]\tLoss: -1957.571680\n",
            "Train Epoch: 72 [8960/60000 (15%)]\tLoss: -2053.773047\n",
            "Train Epoch: 72 [10240/60000 (17%)]\tLoss: -2280.585156\n",
            "Train Epoch: 72 [11520/60000 (19%)]\tLoss: -1994.566406\n",
            "Train Epoch: 72 [12800/60000 (21%)]\tLoss: -2012.937695\n",
            "Train Epoch: 72 [14080/60000 (23%)]\tLoss: -2188.878711\n",
            "Train Epoch: 72 [15360/60000 (26%)]\tLoss: -1957.983789\n",
            "Train Epoch: 72 [16640/60000 (28%)]\tLoss: -1945.295898\n",
            "Train Epoch: 72 [17920/60000 (30%)]\tLoss: -1950.858789\n",
            "Train Epoch: 72 [19200/60000 (32%)]\tLoss: -2083.451953\n",
            "Train Epoch: 72 [20480/60000 (34%)]\tLoss: -1929.063867\n",
            "Train Epoch: 72 [21760/60000 (36%)]\tLoss: -2231.508203\n",
            "Train Epoch: 72 [23040/60000 (38%)]\tLoss: -2122.550000\n",
            "Train Epoch: 72 [24320/60000 (41%)]\tLoss: -2044.156641\n",
            "Train Epoch: 72 [25600/60000 (43%)]\tLoss: -1931.453125\n",
            "Train Epoch: 72 [26880/60000 (45%)]\tLoss: -1945.390625\n",
            "Train Epoch: 72 [28160/60000 (47%)]\tLoss: -2068.903906\n",
            "Train Epoch: 72 [29440/60000 (49%)]\tLoss: -1947.783789\n",
            "Train Epoch: 72 [30720/60000 (51%)]\tLoss: -1948.744336\n",
            "Train Epoch: 72 [32000/60000 (53%)]\tLoss: -2029.458984\n",
            "Train Epoch: 72 [33280/60000 (55%)]\tLoss: -1984.814648\n",
            "Train Epoch: 72 [34560/60000 (58%)]\tLoss: -2061.939453\n",
            "Train Epoch: 72 [35840/60000 (60%)]\tLoss: -1993.610156\n",
            "Train Epoch: 72 [37120/60000 (62%)]\tLoss: -2079.343359\n",
            "Train Epoch: 72 [38400/60000 (64%)]\tLoss: -1968.436719\n",
            "Train Epoch: 72 [39680/60000 (66%)]\tLoss: -1969.500391\n",
            "Train Epoch: 72 [40960/60000 (68%)]\tLoss: -2028.999805\n",
            "Train Epoch: 72 [42240/60000 (70%)]\tLoss: -2007.467773\n",
            "Train Epoch: 72 [43520/60000 (72%)]\tLoss: -1948.515234\n",
            "Train Epoch: 72 [44800/60000 (75%)]\tLoss: -1985.526172\n",
            "Train Epoch: 72 [46080/60000 (77%)]\tLoss: -2005.256641\n",
            "Train Epoch: 72 [47360/60000 (79%)]\tLoss: -2087.796484\n",
            "Train Epoch: 72 [48640/60000 (81%)]\tLoss: -2022.184180\n",
            "Train Epoch: 72 [49920/60000 (83%)]\tLoss: -2014.576172\n",
            "Train Epoch: 72 [51200/60000 (85%)]\tLoss: -2210.177734\n",
            "Train Epoch: 72 [52480/60000 (87%)]\tLoss: -1960.561914\n",
            "Train Epoch: 72 [53760/60000 (90%)]\tLoss: -1966.139844\n",
            "Train Epoch: 72 [55040/60000 (92%)]\tLoss: -1961.756055\n",
            "Train Epoch: 72 [56320/60000 (94%)]\tLoss: -1939.228125\n",
            "Train Epoch: 72 [57600/60000 (96%)]\tLoss: -1958.368359\n",
            "Train Epoch: 72 [58880/60000 (98%)]\tLoss: -1908.488867\n",
            "====> Epoch: 72 Average loss: -20050.9797\n",
            "Train Epoch: 73 [0/60000 (0%)]\tLoss: -19641.619141\n",
            "Train Epoch: 73 [1280/60000 (2%)]\tLoss: -2229.812305\n",
            "Train Epoch: 73 [2560/60000 (4%)]\tLoss: -1931.521094\n",
            "Train Epoch: 73 [3840/60000 (6%)]\tLoss: -1990.316797\n",
            "Train Epoch: 73 [5120/60000 (9%)]\tLoss: -2050.606836\n",
            "Train Epoch: 73 [6400/60000 (11%)]\tLoss: -2004.033594\n",
            "Train Epoch: 73 [7680/60000 (13%)]\tLoss: -2008.813477\n",
            "Train Epoch: 73 [8960/60000 (15%)]\tLoss: -2045.725586\n",
            "Train Epoch: 73 [10240/60000 (17%)]\tLoss: -2256.899805\n",
            "Train Epoch: 73 [11520/60000 (19%)]\tLoss: -1988.794141\n",
            "Train Epoch: 73 [12800/60000 (21%)]\tLoss: -1996.406055\n",
            "Train Epoch: 73 [14080/60000 (23%)]\tLoss: -2195.523828\n",
            "Train Epoch: 73 [15360/60000 (26%)]\tLoss: -1969.400000\n",
            "Train Epoch: 73 [16640/60000 (28%)]\tLoss: -1945.726367\n",
            "Train Epoch: 73 [17920/60000 (30%)]\tLoss: -1964.854102\n",
            "Train Epoch: 73 [19200/60000 (32%)]\tLoss: -2075.679297\n",
            "Train Epoch: 73 [20480/60000 (34%)]\tLoss: -1913.100586\n",
            "Train Epoch: 73 [21760/60000 (36%)]\tLoss: -2213.148828\n",
            "Train Epoch: 73 [23040/60000 (38%)]\tLoss: -2164.641211\n",
            "Train Epoch: 73 [24320/60000 (41%)]\tLoss: -2043.867383\n",
            "Train Epoch: 73 [25600/60000 (43%)]\tLoss: -1937.372070\n",
            "Train Epoch: 73 [26880/60000 (45%)]\tLoss: -1948.180078\n",
            "Train Epoch: 73 [28160/60000 (47%)]\tLoss: -2028.816602\n",
            "Train Epoch: 73 [29440/60000 (49%)]\tLoss: -1974.805078\n",
            "Train Epoch: 73 [30720/60000 (51%)]\tLoss: -1974.864453\n",
            "Train Epoch: 73 [32000/60000 (53%)]\tLoss: -2026.626367\n",
            "Train Epoch: 73 [33280/60000 (55%)]\tLoss: -1964.236133\n",
            "Train Epoch: 73 [34560/60000 (58%)]\tLoss: -2002.094922\n",
            "Train Epoch: 73 [35840/60000 (60%)]\tLoss: -1989.704102\n",
            "Train Epoch: 73 [37120/60000 (62%)]\tLoss: -2057.623047\n",
            "Train Epoch: 73 [38400/60000 (64%)]\tLoss: -1980.718555\n",
            "Train Epoch: 73 [39680/60000 (66%)]\tLoss: -1977.969141\n",
            "Train Epoch: 73 [40960/60000 (68%)]\tLoss: -2079.142578\n",
            "Train Epoch: 73 [42240/60000 (70%)]\tLoss: -2025.696680\n",
            "Train Epoch: 73 [43520/60000 (72%)]\tLoss: -1967.667969\n",
            "Train Epoch: 73 [44800/60000 (75%)]\tLoss: -2005.977930\n",
            "Train Epoch: 73 [46080/60000 (77%)]\tLoss: -2011.929102\n",
            "Train Epoch: 73 [47360/60000 (79%)]\tLoss: -2075.538281\n",
            "Train Epoch: 73 [48640/60000 (81%)]\tLoss: -2030.247266\n",
            "Train Epoch: 73 [49920/60000 (83%)]\tLoss: -2011.656836\n",
            "Train Epoch: 73 [51200/60000 (85%)]\tLoss: -2143.286523\n",
            "Train Epoch: 73 [52480/60000 (87%)]\tLoss: -1975.108398\n",
            "Train Epoch: 73 [53760/60000 (90%)]\tLoss: -1970.056641\n",
            "Train Epoch: 73 [55040/60000 (92%)]\tLoss: -1956.834375\n",
            "Train Epoch: 73 [56320/60000 (94%)]\tLoss: -1939.649805\n",
            "Train Epoch: 73 [57600/60000 (96%)]\tLoss: -1973.463281\n",
            "Train Epoch: 73 [58880/60000 (98%)]\tLoss: -1921.961914\n",
            "====> Epoch: 73 Average loss: -20092.9558\n",
            "Train Epoch: 74 [0/60000 (0%)]\tLoss: -19741.113281\n",
            "Train Epoch: 74 [1280/60000 (2%)]\tLoss: -2212.185742\n",
            "Train Epoch: 74 [2560/60000 (4%)]\tLoss: -1935.007422\n",
            "Train Epoch: 74 [3840/60000 (6%)]\tLoss: -1978.893164\n",
            "Train Epoch: 74 [5120/60000 (9%)]\tLoss: -2117.620313\n",
            "Train Epoch: 74 [6400/60000 (11%)]\tLoss: -1980.681445\n",
            "Train Epoch: 74 [7680/60000 (13%)]\tLoss: -2053.163867\n",
            "Train Epoch: 74 [8960/60000 (15%)]\tLoss: -2072.967578\n",
            "Train Epoch: 74 [10240/60000 (17%)]\tLoss: -2316.649609\n",
            "Train Epoch: 74 [11520/60000 (19%)]\tLoss: -2011.812109\n",
            "Train Epoch: 74 [12800/60000 (21%)]\tLoss: -1991.827734\n",
            "Train Epoch: 74 [14080/60000 (23%)]\tLoss: -2178.013477\n",
            "Train Epoch: 74 [15360/60000 (26%)]\tLoss: -1986.276172\n",
            "Train Epoch: 74 [16640/60000 (28%)]\tLoss: -1945.854687\n",
            "Train Epoch: 74 [17920/60000 (30%)]\tLoss: -1979.358984\n",
            "Train Epoch: 74 [19200/60000 (32%)]\tLoss: -2124.294922\n",
            "Train Epoch: 74 [20480/60000 (34%)]\tLoss: -1929.017188\n",
            "Train Epoch: 74 [21760/60000 (36%)]\tLoss: -2222.202148\n",
            "Train Epoch: 74 [23040/60000 (38%)]\tLoss: -2141.384375\n",
            "Train Epoch: 74 [24320/60000 (41%)]\tLoss: -2035.372852\n",
            "Train Epoch: 74 [25600/60000 (43%)]\tLoss: -1937.136719\n",
            "Train Epoch: 74 [26880/60000 (45%)]\tLoss: -1961.136719\n",
            "Train Epoch: 74 [28160/60000 (47%)]\tLoss: -2010.823437\n",
            "Train Epoch: 74 [29440/60000 (49%)]\tLoss: -1963.880078\n",
            "Train Epoch: 74 [30720/60000 (51%)]\tLoss: -1968.782617\n",
            "Train Epoch: 74 [32000/60000 (53%)]\tLoss: -2043.104687\n",
            "Train Epoch: 74 [33280/60000 (55%)]\tLoss: -1980.091211\n",
            "Train Epoch: 74 [34560/60000 (58%)]\tLoss: -2018.346094\n",
            "Train Epoch: 74 [35840/60000 (60%)]\tLoss: -1983.657422\n",
            "Train Epoch: 74 [37120/60000 (62%)]\tLoss: -2028.941016\n",
            "Train Epoch: 74 [38400/60000 (64%)]\tLoss: -1999.044531\n",
            "Train Epoch: 74 [39680/60000 (66%)]\tLoss: -1977.112109\n",
            "Train Epoch: 74 [40960/60000 (68%)]\tLoss: -2091.196289\n",
            "Train Epoch: 74 [42240/60000 (70%)]\tLoss: -2045.976172\n",
            "Train Epoch: 74 [43520/60000 (72%)]\tLoss: -1963.527148\n",
            "Train Epoch: 74 [44800/60000 (75%)]\tLoss: -2016.640625\n",
            "Train Epoch: 74 [46080/60000 (77%)]\tLoss: -2073.589844\n",
            "Train Epoch: 74 [47360/60000 (79%)]\tLoss: -2119.486133\n",
            "Train Epoch: 74 [48640/60000 (81%)]\tLoss: -2038.599023\n",
            "Train Epoch: 74 [49920/60000 (83%)]\tLoss: -2011.590234\n",
            "Train Epoch: 74 [51200/60000 (85%)]\tLoss: -2158.910937\n",
            "Train Epoch: 74 [52480/60000 (87%)]\tLoss: -1967.364258\n",
            "Train Epoch: 74 [53760/60000 (90%)]\tLoss: -1975.064258\n",
            "Train Epoch: 74 [55040/60000 (92%)]\tLoss: -1961.752734\n",
            "Train Epoch: 74 [56320/60000 (94%)]\tLoss: -1977.548828\n",
            "Train Epoch: 74 [57600/60000 (96%)]\tLoss: -1975.124023\n",
            "Train Epoch: 74 [58880/60000 (98%)]\tLoss: -1923.425000\n",
            "====> Epoch: 74 Average loss: -20170.1748\n",
            "Train Epoch: 75 [0/60000 (0%)]\tLoss: -19983.796875\n",
            "Train Epoch: 75 [1280/60000 (2%)]\tLoss: -2173.149805\n",
            "Train Epoch: 75 [2560/60000 (4%)]\tLoss: -1943.161914\n",
            "Train Epoch: 75 [3840/60000 (6%)]\tLoss: -1982.430078\n",
            "Train Epoch: 75 [5120/60000 (9%)]\tLoss: -2106.732227\n",
            "Train Epoch: 75 [6400/60000 (11%)]\tLoss: -1964.159375\n",
            "Train Epoch: 75 [7680/60000 (13%)]\tLoss: -1993.633203\n",
            "Train Epoch: 75 [8960/60000 (15%)]\tLoss: -2071.121094\n",
            "Train Epoch: 75 [10240/60000 (17%)]\tLoss: -2371.535937\n",
            "Train Epoch: 75 [11520/60000 (19%)]\tLoss: -1959.495312\n",
            "Train Epoch: 75 [12800/60000 (21%)]\tLoss: -1990.546680\n",
            "Train Epoch: 75 [14080/60000 (23%)]\tLoss: -2171.099805\n",
            "Train Epoch: 75 [15360/60000 (26%)]\tLoss: -1975.841016\n",
            "Train Epoch: 75 [16640/60000 (28%)]\tLoss: -1944.673633\n",
            "Train Epoch: 75 [17920/60000 (30%)]\tLoss: -1992.867383\n",
            "Train Epoch: 75 [19200/60000 (32%)]\tLoss: -2075.776367\n",
            "Train Epoch: 75 [20480/60000 (34%)]\tLoss: -1949.132227\n",
            "Train Epoch: 75 [21760/60000 (36%)]\tLoss: -2235.847070\n",
            "Train Epoch: 75 [23040/60000 (38%)]\tLoss: -2135.552344\n",
            "Train Epoch: 75 [24320/60000 (41%)]\tLoss: -2032.297266\n",
            "Train Epoch: 75 [25600/60000 (43%)]\tLoss: -1952.130273\n",
            "Train Epoch: 75 [26880/60000 (45%)]\tLoss: -1976.117578\n",
            "Train Epoch: 75 [28160/60000 (47%)]\tLoss: -2023.612695\n",
            "Train Epoch: 75 [29440/60000 (49%)]\tLoss: -1980.049023\n",
            "Train Epoch: 75 [30720/60000 (51%)]\tLoss: -1993.974609\n",
            "Train Epoch: 75 [32000/60000 (53%)]\tLoss: -2065.968750\n",
            "Train Epoch: 75 [33280/60000 (55%)]\tLoss: -1957.710938\n",
            "Train Epoch: 75 [34560/60000 (58%)]\tLoss: -1977.482422\n",
            "Train Epoch: 75 [35840/60000 (60%)]\tLoss: -1991.246680\n",
            "Train Epoch: 75 [37120/60000 (62%)]\tLoss: -2128.409180\n",
            "Train Epoch: 75 [38400/60000 (64%)]\tLoss: -2001.322266\n",
            "Train Epoch: 75 [39680/60000 (66%)]\tLoss: -1975.459375\n",
            "Train Epoch: 75 [40960/60000 (68%)]\tLoss: -2077.837305\n",
            "Train Epoch: 75 [42240/60000 (70%)]\tLoss: -2012.779883\n",
            "Train Epoch: 75 [43520/60000 (72%)]\tLoss: -1975.645508\n",
            "Train Epoch: 75 [44800/60000 (75%)]\tLoss: -2015.032227\n",
            "Train Epoch: 75 [46080/60000 (77%)]\tLoss: -2115.175977\n",
            "Train Epoch: 75 [47360/60000 (79%)]\tLoss: -2171.388281\n",
            "Train Epoch: 75 [48640/60000 (81%)]\tLoss: -2038.189258\n",
            "Train Epoch: 75 [49920/60000 (83%)]\tLoss: -2027.286523\n",
            "Train Epoch: 75 [51200/60000 (85%)]\tLoss: -2146.459766\n",
            "Train Epoch: 75 [52480/60000 (87%)]\tLoss: -1941.484766\n",
            "Train Epoch: 75 [53760/60000 (90%)]\tLoss: -1982.153711\n",
            "Train Epoch: 75 [55040/60000 (92%)]\tLoss: -1979.501758\n",
            "Train Epoch: 75 [56320/60000 (94%)]\tLoss: -1976.036523\n",
            "Train Epoch: 75 [57600/60000 (96%)]\tLoss: -2009.304492\n",
            "Train Epoch: 75 [58880/60000 (98%)]\tLoss: -1921.211133\n",
            "====> Epoch: 75 Average loss: -20225.5019\n",
            "Train Epoch: 76 [0/60000 (0%)]\tLoss: -19850.576172\n",
            "Train Epoch: 76 [1280/60000 (2%)]\tLoss: -2200.803125\n",
            "Train Epoch: 76 [2560/60000 (4%)]\tLoss: -1931.824219\n",
            "Train Epoch: 76 [3840/60000 (6%)]\tLoss: -1975.806641\n",
            "Train Epoch: 76 [5120/60000 (9%)]\tLoss: -2090.246875\n",
            "Train Epoch: 76 [6400/60000 (11%)]\tLoss: -1983.206836\n",
            "Train Epoch: 76 [7680/60000 (13%)]\tLoss: -2042.602344\n",
            "Train Epoch: 76 [8960/60000 (15%)]\tLoss: -2056.341406\n",
            "Train Epoch: 76 [10240/60000 (17%)]\tLoss: -2411.879297\n",
            "Train Epoch: 76 [11520/60000 (19%)]\tLoss: -1987.276367\n",
            "Train Epoch: 76 [12800/60000 (21%)]\tLoss: -2005.237695\n",
            "Train Epoch: 76 [14080/60000 (23%)]\tLoss: -2224.359180\n",
            "Train Epoch: 76 [15360/60000 (26%)]\tLoss: -1964.581836\n",
            "Train Epoch: 76 [16640/60000 (28%)]\tLoss: -1940.775391\n",
            "Train Epoch: 76 [17920/60000 (30%)]\tLoss: -2008.809570\n",
            "Train Epoch: 76 [19200/60000 (32%)]\tLoss: -2106.716602\n",
            "Train Epoch: 76 [20480/60000 (34%)]\tLoss: -1930.838672\n",
            "Train Epoch: 76 [21760/60000 (36%)]\tLoss: -2275.724414\n",
            "Train Epoch: 76 [23040/60000 (38%)]\tLoss: -2165.618750\n",
            "Train Epoch: 76 [24320/60000 (41%)]\tLoss: -2094.403906\n",
            "Train Epoch: 76 [25600/60000 (43%)]\tLoss: -1938.941992\n",
            "Train Epoch: 76 [26880/60000 (45%)]\tLoss: -1969.948047\n",
            "Train Epoch: 76 [28160/60000 (47%)]\tLoss: -2000.725781\n",
            "Train Epoch: 76 [29440/60000 (49%)]\tLoss: -1954.323633\n",
            "Train Epoch: 76 [30720/60000 (51%)]\tLoss: -1995.402734\n",
            "Train Epoch: 76 [32000/60000 (53%)]\tLoss: -2032.775586\n",
            "Train Epoch: 76 [33280/60000 (55%)]\tLoss: -1960.880469\n",
            "Train Epoch: 76 [34560/60000 (58%)]\tLoss: -1993.460547\n",
            "Train Epoch: 76 [35840/60000 (60%)]\tLoss: -1989.232227\n",
            "Train Epoch: 76 [37120/60000 (62%)]\tLoss: -2043.570703\n",
            "Train Epoch: 76 [38400/60000 (64%)]\tLoss: -1998.226953\n",
            "Train Epoch: 76 [39680/60000 (66%)]\tLoss: -1985.571094\n",
            "Train Epoch: 76 [40960/60000 (68%)]\tLoss: -2080.378906\n",
            "Train Epoch: 76 [42240/60000 (70%)]\tLoss: -1981.494141\n",
            "Train Epoch: 76 [43520/60000 (72%)]\tLoss: -1949.495898\n",
            "Train Epoch: 76 [44800/60000 (75%)]\tLoss: -2001.434766\n",
            "Train Epoch: 76 [46080/60000 (77%)]\tLoss: -2142.694336\n",
            "Train Epoch: 76 [47360/60000 (79%)]\tLoss: -2137.763086\n",
            "Train Epoch: 76 [48640/60000 (81%)]\tLoss: -2091.964648\n",
            "Train Epoch: 76 [49920/60000 (83%)]\tLoss: -2031.177734\n",
            "Train Epoch: 76 [51200/60000 (85%)]\tLoss: -2132.706055\n",
            "Train Epoch: 76 [52480/60000 (87%)]\tLoss: -1920.964844\n",
            "Train Epoch: 76 [53760/60000 (90%)]\tLoss: -1991.211133\n",
            "Train Epoch: 76 [55040/60000 (92%)]\tLoss: -1982.222266\n",
            "Train Epoch: 76 [56320/60000 (94%)]\tLoss: -1988.672852\n",
            "Train Epoch: 76 [57600/60000 (96%)]\tLoss: -1999.432617\n",
            "Train Epoch: 76 [58880/60000 (98%)]\tLoss: -1938.840039\n",
            "====> Epoch: 76 Average loss: -20257.6602\n",
            "Train Epoch: 77 [0/60000 (0%)]\tLoss: -19753.275391\n",
            "Train Epoch: 77 [1280/60000 (2%)]\tLoss: -2197.704102\n",
            "Train Epoch: 77 [2560/60000 (4%)]\tLoss: -1935.980664\n",
            "Train Epoch: 77 [3840/60000 (6%)]\tLoss: -1976.920312\n",
            "Train Epoch: 77 [5120/60000 (9%)]\tLoss: -2085.920508\n",
            "Train Epoch: 77 [6400/60000 (11%)]\tLoss: -1966.907813\n",
            "Train Epoch: 77 [7680/60000 (13%)]\tLoss: -2004.843555\n",
            "Train Epoch: 77 [8960/60000 (15%)]\tLoss: -2093.097852\n",
            "Train Epoch: 77 [10240/60000 (17%)]\tLoss: -2409.991016\n",
            "Train Epoch: 77 [11520/60000 (19%)]\tLoss: -2058.414844\n",
            "Train Epoch: 77 [12800/60000 (21%)]\tLoss: -2030.272461\n",
            "Train Epoch: 77 [14080/60000 (23%)]\tLoss: -2242.400000\n",
            "Train Epoch: 77 [15360/60000 (26%)]\tLoss: -1989.622266\n",
            "Train Epoch: 77 [16640/60000 (28%)]\tLoss: -1939.538672\n",
            "Train Epoch: 77 [17920/60000 (30%)]\tLoss: -1998.221875\n",
            "Train Epoch: 77 [19200/60000 (32%)]\tLoss: -2145.504102\n",
            "Train Epoch: 77 [20480/60000 (34%)]\tLoss: -1944.862695\n",
            "Train Epoch: 77 [21760/60000 (36%)]\tLoss: -2293.888867\n",
            "Train Epoch: 77 [23040/60000 (38%)]\tLoss: -2146.427344\n",
            "Train Epoch: 77 [24320/60000 (41%)]\tLoss: -2075.003320\n",
            "Train Epoch: 77 [25600/60000 (43%)]\tLoss: -1937.265625\n",
            "Train Epoch: 77 [26880/60000 (45%)]\tLoss: -1971.592383\n",
            "Train Epoch: 77 [28160/60000 (47%)]\tLoss: -2025.456836\n",
            "Train Epoch: 77 [29440/60000 (49%)]\tLoss: -1968.399805\n",
            "Train Epoch: 77 [30720/60000 (51%)]\tLoss: -1967.516016\n",
            "Train Epoch: 77 [32000/60000 (53%)]\tLoss: -2072.453125\n",
            "Train Epoch: 77 [33280/60000 (55%)]\tLoss: -1976.716406\n",
            "Train Epoch: 77 [34560/60000 (58%)]\tLoss: -2020.261133\n",
            "Train Epoch: 77 [35840/60000 (60%)]\tLoss: -1987.126563\n",
            "Train Epoch: 77 [37120/60000 (62%)]\tLoss: -2037.851172\n",
            "Train Epoch: 77 [38400/60000 (64%)]\tLoss: -1994.218945\n",
            "Train Epoch: 77 [39680/60000 (66%)]\tLoss: -1992.808008\n",
            "Train Epoch: 77 [40960/60000 (68%)]\tLoss: -2085.345898\n",
            "Train Epoch: 77 [42240/60000 (70%)]\tLoss: -2008.252930\n",
            "Train Epoch: 77 [43520/60000 (72%)]\tLoss: -1948.487305\n",
            "Train Epoch: 77 [44800/60000 (75%)]\tLoss: -2002.012305\n",
            "Train Epoch: 77 [46080/60000 (77%)]\tLoss: -2110.789258\n",
            "Train Epoch: 77 [47360/60000 (79%)]\tLoss: -2142.060937\n",
            "Train Epoch: 77 [48640/60000 (81%)]\tLoss: -2121.531250\n",
            "Train Epoch: 77 [49920/60000 (83%)]\tLoss: -2034.263867\n",
            "Train Epoch: 77 [51200/60000 (85%)]\tLoss: -2174.949023\n",
            "Train Epoch: 77 [52480/60000 (87%)]\tLoss: -1938.976367\n",
            "Train Epoch: 77 [53760/60000 (90%)]\tLoss: -2015.820508\n",
            "Train Epoch: 77 [55040/60000 (92%)]\tLoss: -1992.031445\n",
            "Train Epoch: 77 [56320/60000 (94%)]\tLoss: -1962.424805\n",
            "Train Epoch: 77 [57600/60000 (96%)]\tLoss: -1990.220313\n",
            "Train Epoch: 77 [58880/60000 (98%)]\tLoss: -1946.573437\n",
            "====> Epoch: 77 Average loss: -20310.8418\n",
            "Train Epoch: 78 [0/60000 (0%)]\tLoss: -19841.384766\n",
            "Train Epoch: 78 [1280/60000 (2%)]\tLoss: -2191.894531\n",
            "Train Epoch: 78 [2560/60000 (4%)]\tLoss: -1950.075000\n",
            "Train Epoch: 78 [3840/60000 (6%)]\tLoss: -2015.184375\n",
            "Train Epoch: 78 [5120/60000 (9%)]\tLoss: -2061.317383\n",
            "Train Epoch: 78 [6400/60000 (11%)]\tLoss: -1971.965039\n",
            "Train Epoch: 78 [7680/60000 (13%)]\tLoss: -1981.780664\n",
            "Train Epoch: 78 [8960/60000 (15%)]\tLoss: -2077.150781\n",
            "Train Epoch: 78 [10240/60000 (17%)]\tLoss: -2437.591211\n",
            "Train Epoch: 78 [11520/60000 (19%)]\tLoss: -2043.079883\n",
            "Train Epoch: 78 [12800/60000 (21%)]\tLoss: -2036.877148\n",
            "Train Epoch: 78 [14080/60000 (23%)]\tLoss: -2225.501367\n",
            "Train Epoch: 78 [15360/60000 (26%)]\tLoss: -1958.736523\n",
            "Train Epoch: 78 [16640/60000 (28%)]\tLoss: -1937.758594\n",
            "Train Epoch: 78 [17920/60000 (30%)]\tLoss: -1998.258789\n",
            "Train Epoch: 78 [19200/60000 (32%)]\tLoss: -2149.320898\n",
            "Train Epoch: 78 [20480/60000 (34%)]\tLoss: -1942.600977\n",
            "Train Epoch: 78 [21760/60000 (36%)]\tLoss: -2240.138281\n",
            "Train Epoch: 78 [23040/60000 (38%)]\tLoss: -2160.125781\n",
            "Train Epoch: 78 [24320/60000 (41%)]\tLoss: -2088.536328\n",
            "Train Epoch: 78 [25600/60000 (43%)]\tLoss: -1968.883008\n",
            "Train Epoch: 78 [26880/60000 (45%)]\tLoss: -1992.612695\n",
            "Train Epoch: 78 [28160/60000 (47%)]\tLoss: -2078.006055\n",
            "Train Epoch: 78 [29440/60000 (49%)]\tLoss: -1954.118750\n",
            "Train Epoch: 78 [30720/60000 (51%)]\tLoss: -1950.370508\n",
            "Train Epoch: 78 [32000/60000 (53%)]\tLoss: -2025.662695\n",
            "Train Epoch: 78 [33280/60000 (55%)]\tLoss: -1941.420898\n",
            "Train Epoch: 78 [34560/60000 (58%)]\tLoss: -2024.949023\n",
            "Train Epoch: 78 [35840/60000 (60%)]\tLoss: -1995.406055\n",
            "Train Epoch: 78 [37120/60000 (62%)]\tLoss: -2064.889258\n",
            "Train Epoch: 78 [38400/60000 (64%)]\tLoss: -1987.445312\n",
            "Train Epoch: 78 [39680/60000 (66%)]\tLoss: -1991.672070\n",
            "Train Epoch: 78 [40960/60000 (68%)]\tLoss: -2055.072266\n",
            "Train Epoch: 78 [42240/60000 (70%)]\tLoss: -2063.257617\n",
            "Train Epoch: 78 [43520/60000 (72%)]\tLoss: -1992.988672\n",
            "Train Epoch: 78 [44800/60000 (75%)]\tLoss: -2020.724609\n",
            "Train Epoch: 78 [46080/60000 (77%)]\tLoss: -2053.212891\n",
            "Train Epoch: 78 [47360/60000 (79%)]\tLoss: -2144.318750\n",
            "Train Epoch: 78 [48640/60000 (81%)]\tLoss: -2084.357227\n",
            "Train Epoch: 78 [49920/60000 (83%)]\tLoss: -2030.956055\n",
            "Train Epoch: 78 [51200/60000 (85%)]\tLoss: -2186.793555\n",
            "Train Epoch: 78 [52480/60000 (87%)]\tLoss: -1940.488672\n",
            "Train Epoch: 78 [53760/60000 (90%)]\tLoss: -1989.038086\n",
            "Train Epoch: 78 [55040/60000 (92%)]\tLoss: -1976.527539\n",
            "Train Epoch: 78 [56320/60000 (94%)]\tLoss: -1955.745312\n",
            "Train Epoch: 78 [57600/60000 (96%)]\tLoss: -1990.119727\n",
            "Train Epoch: 78 [58880/60000 (98%)]\tLoss: -1968.943555\n",
            "====> Epoch: 78 Average loss: -20345.2557\n",
            "Train Epoch: 79 [0/60000 (0%)]\tLoss: -20044.552734\n",
            "Train Epoch: 79 [1280/60000 (2%)]\tLoss: -2196.148438\n",
            "Train Epoch: 79 [2560/60000 (4%)]\tLoss: -1952.877734\n",
            "Train Epoch: 79 [3840/60000 (6%)]\tLoss: -1978.594531\n",
            "Train Epoch: 79 [5120/60000 (9%)]\tLoss: -2109.528320\n",
            "Train Epoch: 79 [6400/60000 (11%)]\tLoss: -2004.175391\n",
            "Train Epoch: 79 [7680/60000 (13%)]\tLoss: -1986.466797\n",
            "Train Epoch: 79 [8960/60000 (15%)]\tLoss: -2080.041406\n",
            "Train Epoch: 79 [10240/60000 (17%)]\tLoss: -2443.654297\n",
            "Train Epoch: 79 [11520/60000 (19%)]\tLoss: -2019.668750\n",
            "Train Epoch: 79 [12800/60000 (21%)]\tLoss: -2030.875391\n",
            "Train Epoch: 79 [14080/60000 (23%)]\tLoss: -2258.218750\n",
            "Train Epoch: 79 [15360/60000 (26%)]\tLoss: -1978.800195\n",
            "Train Epoch: 79 [16640/60000 (28%)]\tLoss: -1943.630859\n",
            "Train Epoch: 79 [17920/60000 (30%)]\tLoss: -2001.269336\n",
            "Train Epoch: 79 [19200/60000 (32%)]\tLoss: -2136.501953\n",
            "Train Epoch: 79 [20480/60000 (34%)]\tLoss: -1965.920898\n",
            "Train Epoch: 79 [21760/60000 (36%)]\tLoss: -2293.878906\n",
            "Train Epoch: 79 [23040/60000 (38%)]\tLoss: -2157.004297\n",
            "Train Epoch: 79 [24320/60000 (41%)]\tLoss: -2075.702734\n",
            "Train Epoch: 79 [25600/60000 (43%)]\tLoss: -1963.321289\n",
            "Train Epoch: 79 [26880/60000 (45%)]\tLoss: -1981.348633\n",
            "Train Epoch: 79 [28160/60000 (47%)]\tLoss: -2058.747266\n",
            "Train Epoch: 79 [29440/60000 (49%)]\tLoss: -1967.629688\n",
            "Train Epoch: 79 [30720/60000 (51%)]\tLoss: -1972.406250\n",
            "Train Epoch: 79 [32000/60000 (53%)]\tLoss: -2039.539258\n",
            "Train Epoch: 79 [33280/60000 (55%)]\tLoss: -1972.038086\n",
            "Train Epoch: 79 [34560/60000 (58%)]\tLoss: -2059.687109\n",
            "Train Epoch: 79 [35840/60000 (60%)]\tLoss: -1998.463281\n",
            "Train Epoch: 79 [37120/60000 (62%)]\tLoss: -2105.485156\n",
            "Train Epoch: 79 [38400/60000 (64%)]\tLoss: -1982.328906\n",
            "Train Epoch: 79 [39680/60000 (66%)]\tLoss: -1978.910547\n",
            "Train Epoch: 79 [40960/60000 (68%)]\tLoss: -2056.742383\n",
            "Train Epoch: 79 [42240/60000 (70%)]\tLoss: -2009.666016\n",
            "Train Epoch: 79 [43520/60000 (72%)]\tLoss: -1984.946680\n",
            "Train Epoch: 79 [44800/60000 (75%)]\tLoss: -2003.007031\n",
            "Train Epoch: 79 [46080/60000 (77%)]\tLoss: -2044.083789\n",
            "Train Epoch: 79 [47360/60000 (79%)]\tLoss: -2096.315430\n",
            "Train Epoch: 79 [48640/60000 (81%)]\tLoss: -2066.082617\n",
            "Train Epoch: 79 [49920/60000 (83%)]\tLoss: -2018.523438\n",
            "Train Epoch: 79 [51200/60000 (85%)]\tLoss: -2127.815430\n",
            "Train Epoch: 79 [52480/60000 (87%)]\tLoss: -1952.816211\n",
            "Train Epoch: 79 [53760/60000 (90%)]\tLoss: -1967.592187\n",
            "Train Epoch: 79 [55040/60000 (92%)]\tLoss: -1989.383594\n",
            "Train Epoch: 79 [56320/60000 (94%)]\tLoss: -1958.022266\n",
            "Train Epoch: 79 [57600/60000 (96%)]\tLoss: -2003.506055\n",
            "Train Epoch: 79 [58880/60000 (98%)]\tLoss: -1946.935742\n",
            "====> Epoch: 79 Average loss: -20325.8424\n",
            "Train Epoch: 80 [0/60000 (0%)]\tLoss: -19864.609375\n",
            "Train Epoch: 80 [1280/60000 (2%)]\tLoss: -2204.720313\n",
            "Train Epoch: 80 [2560/60000 (4%)]\tLoss: -1938.067383\n",
            "Train Epoch: 80 [3840/60000 (6%)]\tLoss: -1987.622656\n",
            "Train Epoch: 80 [5120/60000 (9%)]\tLoss: -2115.860547\n",
            "Train Epoch: 80 [6400/60000 (11%)]\tLoss: -1965.231836\n",
            "Train Epoch: 80 [7680/60000 (13%)]\tLoss: -1979.824414\n",
            "Train Epoch: 80 [8960/60000 (15%)]\tLoss: -2076.641602\n",
            "Train Epoch: 80 [10240/60000 (17%)]\tLoss: -2420.105859\n",
            "Train Epoch: 80 [11520/60000 (19%)]\tLoss: -2005.160742\n",
            "Train Epoch: 80 [12800/60000 (21%)]\tLoss: -2014.569922\n",
            "Train Epoch: 80 [14080/60000 (23%)]\tLoss: -2218.753906\n",
            "Train Epoch: 80 [15360/60000 (26%)]\tLoss: -1979.350586\n",
            "Train Epoch: 80 [16640/60000 (28%)]\tLoss: -1943.340234\n",
            "Train Epoch: 80 [17920/60000 (30%)]\tLoss: -1989.534570\n",
            "Train Epoch: 80 [19200/60000 (32%)]\tLoss: -2142.965039\n",
            "Train Epoch: 80 [20480/60000 (34%)]\tLoss: -1927.583398\n",
            "Train Epoch: 80 [21760/60000 (36%)]\tLoss: -2246.778320\n",
            "Train Epoch: 80 [23040/60000 (38%)]\tLoss: -2138.293945\n",
            "Train Epoch: 80 [24320/60000 (41%)]\tLoss: -2079.764453\n",
            "Train Epoch: 80 [25600/60000 (43%)]\tLoss: -1985.450391\n",
            "Train Epoch: 80 [26880/60000 (45%)]\tLoss: -1982.611133\n",
            "Train Epoch: 80 [28160/60000 (47%)]\tLoss: -2030.281445\n",
            "Train Epoch: 80 [29440/60000 (49%)]\tLoss: -1952.666016\n",
            "Train Epoch: 80 [30720/60000 (51%)]\tLoss: -1958.618750\n",
            "Train Epoch: 80 [32000/60000 (53%)]\tLoss: -2043.810156\n",
            "Train Epoch: 80 [33280/60000 (55%)]\tLoss: -1993.544922\n",
            "Train Epoch: 80 [34560/60000 (58%)]\tLoss: -2063.707812\n",
            "Train Epoch: 80 [35840/60000 (60%)]\tLoss: -1992.584766\n",
            "Train Epoch: 80 [37120/60000 (62%)]\tLoss: -2111.461719\n",
            "Train Epoch: 80 [38400/60000 (64%)]\tLoss: -1972.257617\n",
            "Train Epoch: 80 [39680/60000 (66%)]\tLoss: -1975.355859\n",
            "Train Epoch: 80 [40960/60000 (68%)]\tLoss: -2048.512891\n",
            "Train Epoch: 80 [42240/60000 (70%)]\tLoss: -2018.699609\n",
            "Train Epoch: 80 [43520/60000 (72%)]\tLoss: -1984.281250\n",
            "Train Epoch: 80 [44800/60000 (75%)]\tLoss: -2004.814063\n",
            "Train Epoch: 80 [46080/60000 (77%)]\tLoss: -2057.494141\n",
            "Train Epoch: 80 [47360/60000 (79%)]\tLoss: -2114.678906\n",
            "Train Epoch: 80 [48640/60000 (81%)]\tLoss: -2092.341602\n",
            "Train Epoch: 80 [49920/60000 (83%)]\tLoss: -2033.343164\n",
            "Train Epoch: 80 [51200/60000 (85%)]\tLoss: -2181.204102\n",
            "Train Epoch: 80 [52480/60000 (87%)]\tLoss: -1937.638867\n",
            "Train Epoch: 80 [53760/60000 (90%)]\tLoss: -1989.871094\n",
            "Train Epoch: 80 [55040/60000 (92%)]\tLoss: -1975.914062\n",
            "Train Epoch: 80 [56320/60000 (94%)]\tLoss: -1948.758789\n",
            "Train Epoch: 80 [57600/60000 (96%)]\tLoss: -1993.178125\n",
            "Train Epoch: 80 [58880/60000 (98%)]\tLoss: -1959.737109\n",
            "====> Epoch: 80 Average loss: -20304.5236\n",
            "Train Epoch: 81 [0/60000 (0%)]\tLoss: -19913.662109\n",
            "Train Epoch: 81 [1280/60000 (2%)]\tLoss: -2208.879687\n",
            "Train Epoch: 81 [2560/60000 (4%)]\tLoss: -1956.161328\n",
            "Train Epoch: 81 [3840/60000 (6%)]\tLoss: -1995.757227\n",
            "Train Epoch: 81 [5120/60000 (9%)]\tLoss: -2125.218945\n",
            "Train Epoch: 81 [6400/60000 (11%)]\tLoss: -1978.837891\n",
            "Train Epoch: 81 [7680/60000 (13%)]\tLoss: -2005.862695\n",
            "Train Epoch: 81 [8960/60000 (15%)]\tLoss: -2097.326367\n",
            "Train Epoch: 81 [10240/60000 (17%)]\tLoss: -2394.417383\n",
            "Train Epoch: 81 [11520/60000 (19%)]\tLoss: -2020.474414\n",
            "Train Epoch: 81 [12800/60000 (21%)]\tLoss: -2037.679297\n",
            "Train Epoch: 81 [14080/60000 (23%)]\tLoss: -2274.362305\n",
            "Train Epoch: 81 [15360/60000 (26%)]\tLoss: -1960.535352\n",
            "Train Epoch: 81 [16640/60000 (28%)]\tLoss: -1945.351953\n",
            "Train Epoch: 81 [17920/60000 (30%)]\tLoss: -2007.138672\n",
            "Train Epoch: 81 [19200/60000 (32%)]\tLoss: -2131.660742\n",
            "Train Epoch: 81 [20480/60000 (34%)]\tLoss: -1951.854687\n",
            "Train Epoch: 81 [21760/60000 (36%)]\tLoss: -2292.211523\n",
            "Train Epoch: 81 [23040/60000 (38%)]\tLoss: -2168.178516\n",
            "Train Epoch: 81 [24320/60000 (41%)]\tLoss: -2060.451172\n",
            "Train Epoch: 81 [25600/60000 (43%)]\tLoss: -1942.861523\n",
            "Train Epoch: 81 [26880/60000 (45%)]\tLoss: -1980.226953\n",
            "Train Epoch: 81 [28160/60000 (47%)]\tLoss: -2047.953906\n",
            "Train Epoch: 81 [29440/60000 (49%)]\tLoss: -1948.749609\n",
            "Train Epoch: 81 [30720/60000 (51%)]\tLoss: -1953.490625\n",
            "Train Epoch: 81 [32000/60000 (53%)]\tLoss: -2037.185742\n",
            "Train Epoch: 81 [33280/60000 (55%)]\tLoss: -1983.157227\n",
            "Train Epoch: 81 [34560/60000 (58%)]\tLoss: -2026.762695\n",
            "Train Epoch: 81 [35840/60000 (60%)]\tLoss: -1995.414844\n",
            "Train Epoch: 81 [37120/60000 (62%)]\tLoss: -2044.119336\n",
            "Train Epoch: 81 [38400/60000 (64%)]\tLoss: -1973.158594\n",
            "Train Epoch: 81 [39680/60000 (66%)]\tLoss: -1969.721875\n",
            "Train Epoch: 81 [40960/60000 (68%)]\tLoss: -2068.327734\n",
            "Train Epoch: 81 [42240/60000 (70%)]\tLoss: -2029.312500\n",
            "Train Epoch: 81 [43520/60000 (72%)]\tLoss: -1952.204297\n",
            "Train Epoch: 81 [44800/60000 (75%)]\tLoss: -2010.112695\n",
            "Train Epoch: 81 [46080/60000 (77%)]\tLoss: -2068.198828\n",
            "Train Epoch: 81 [47360/60000 (79%)]\tLoss: -2118.083203\n",
            "Train Epoch: 81 [48640/60000 (81%)]\tLoss: -2069.650977\n",
            "Train Epoch: 81 [49920/60000 (83%)]\tLoss: -2019.367383\n",
            "Train Epoch: 81 [51200/60000 (85%)]\tLoss: -2184.582227\n",
            "Train Epoch: 81 [52480/60000 (87%)]\tLoss: -1947.387500\n",
            "Train Epoch: 81 [53760/60000 (90%)]\tLoss: -1997.011914\n",
            "Train Epoch: 81 [55040/60000 (92%)]\tLoss: -1965.227148\n",
            "Train Epoch: 81 [56320/60000 (94%)]\tLoss: -1958.837500\n",
            "Train Epoch: 81 [57600/60000 (96%)]\tLoss: -2019.857227\n",
            "Train Epoch: 81 [58880/60000 (98%)]\tLoss: -1917.666797\n",
            "====> Epoch: 81 Average loss: -20302.2690\n",
            "Train Epoch: 82 [0/60000 (0%)]\tLoss: -19869.666016\n",
            "Train Epoch: 82 [1280/60000 (2%)]\tLoss: -2207.404687\n",
            "Train Epoch: 82 [2560/60000 (4%)]\tLoss: -1950.563672\n",
            "Train Epoch: 82 [3840/60000 (6%)]\tLoss: -1977.776172\n",
            "Train Epoch: 82 [5120/60000 (9%)]\tLoss: -2064.043164\n",
            "Train Epoch: 82 [6400/60000 (11%)]\tLoss: -1998.697852\n",
            "Train Epoch: 82 [7680/60000 (13%)]\tLoss: -1989.099023\n",
            "Train Epoch: 82 [8960/60000 (15%)]\tLoss: -2068.493750\n",
            "Train Epoch: 82 [10240/60000 (17%)]\tLoss: -2387.771484\n",
            "Train Epoch: 82 [11520/60000 (19%)]\tLoss: -2027.899805\n",
            "Train Epoch: 82 [12800/60000 (21%)]\tLoss: -2027.993359\n",
            "Train Epoch: 82 [14080/60000 (23%)]\tLoss: -2197.001367\n",
            "Train Epoch: 82 [15360/60000 (26%)]\tLoss: -1990.055664\n",
            "Train Epoch: 82 [16640/60000 (28%)]\tLoss: -1946.436328\n",
            "Train Epoch: 82 [17920/60000 (30%)]\tLoss: -2009.633398\n",
            "Train Epoch: 82 [19200/60000 (32%)]\tLoss: -2138.673242\n",
            "Train Epoch: 82 [20480/60000 (34%)]\tLoss: -1918.988867\n",
            "Train Epoch: 82 [21760/60000 (36%)]\tLoss: -2277.483594\n",
            "Train Epoch: 82 [23040/60000 (38%)]\tLoss: -2134.713867\n",
            "Train Epoch: 82 [24320/60000 (41%)]\tLoss: -2033.611328\n",
            "Train Epoch: 82 [25600/60000 (43%)]\tLoss: -1968.773438\n",
            "Train Epoch: 82 [26880/60000 (45%)]\tLoss: -1966.802930\n",
            "Train Epoch: 82 [28160/60000 (47%)]\tLoss: -2030.982227\n",
            "Train Epoch: 82 [29440/60000 (49%)]\tLoss: -1963.542383\n",
            "Train Epoch: 82 [30720/60000 (51%)]\tLoss: -1948.811523\n",
            "Train Epoch: 82 [32000/60000 (53%)]\tLoss: -2029.817773\n",
            "Train Epoch: 82 [33280/60000 (55%)]\tLoss: -2013.852148\n",
            "Train Epoch: 82 [34560/60000 (58%)]\tLoss: -2011.551953\n",
            "Train Epoch: 82 [35840/60000 (60%)]\tLoss: -2002.396094\n",
            "Train Epoch: 82 [37120/60000 (62%)]\tLoss: -2098.513672\n",
            "Train Epoch: 82 [38400/60000 (64%)]\tLoss: -1949.069531\n",
            "Train Epoch: 82 [39680/60000 (66%)]\tLoss: -1973.728125\n",
            "Train Epoch: 82 [40960/60000 (68%)]\tLoss: -2069.146094\n",
            "Train Epoch: 82 [42240/60000 (70%)]\tLoss: -1997.943359\n",
            "Train Epoch: 82 [43520/60000 (72%)]\tLoss: -1972.306250\n",
            "Train Epoch: 82 [44800/60000 (75%)]\tLoss: -2012.757617\n",
            "Train Epoch: 82 [46080/60000 (77%)]\tLoss: -2040.139844\n",
            "Train Epoch: 82 [47360/60000 (79%)]\tLoss: -2102.068164\n",
            "Train Epoch: 82 [48640/60000 (81%)]\tLoss: -2052.987500\n",
            "Train Epoch: 82 [49920/60000 (83%)]\tLoss: -2020.756055\n",
            "Train Epoch: 82 [51200/60000 (85%)]\tLoss: -2166.508789\n",
            "Train Epoch: 82 [52480/60000 (87%)]\tLoss: -1947.620117\n",
            "Train Epoch: 82 [53760/60000 (90%)]\tLoss: -1976.085156\n",
            "Train Epoch: 82 [55040/60000 (92%)]\tLoss: -1982.671680\n",
            "Train Epoch: 82 [56320/60000 (94%)]\tLoss: -1945.691406\n",
            "Train Epoch: 82 [57600/60000 (96%)]\tLoss: -1976.689063\n",
            "Train Epoch: 82 [58880/60000 (98%)]\tLoss: -1962.585938\n",
            "====> Epoch: 82 Average loss: -20278.2397\n",
            "Train Epoch: 83 [0/60000 (0%)]\tLoss: -19846.103516\n",
            "Train Epoch: 83 [1280/60000 (2%)]\tLoss: -2204.821680\n",
            "Train Epoch: 83 [2560/60000 (4%)]\tLoss: -1952.414844\n",
            "Train Epoch: 83 [3840/60000 (6%)]\tLoss: -1981.990430\n",
            "Train Epoch: 83 [5120/60000 (9%)]\tLoss: -2070.390820\n",
            "Train Epoch: 83 [6400/60000 (11%)]\tLoss: -2019.631055\n",
            "Train Epoch: 83 [7680/60000 (13%)]\tLoss: -1959.319531\n",
            "Train Epoch: 83 [8960/60000 (15%)]\tLoss: -2024.041211\n",
            "Train Epoch: 83 [10240/60000 (17%)]\tLoss: -2330.426758\n",
            "Train Epoch: 83 [11520/60000 (19%)]\tLoss: -2009.504883\n",
            "Train Epoch: 83 [12800/60000 (21%)]\tLoss: -2022.364648\n",
            "Train Epoch: 83 [14080/60000 (23%)]\tLoss: -2196.787891\n",
            "Train Epoch: 83 [15360/60000 (26%)]\tLoss: -1996.259375\n",
            "Train Epoch: 83 [16640/60000 (28%)]\tLoss: -1945.364453\n",
            "Train Epoch: 83 [17920/60000 (30%)]\tLoss: -1964.795117\n",
            "Train Epoch: 83 [19200/60000 (32%)]\tLoss: -2144.384180\n",
            "Train Epoch: 83 [20480/60000 (34%)]\tLoss: -1940.537305\n",
            "Train Epoch: 83 [21760/60000 (36%)]\tLoss: -2284.922461\n",
            "Train Epoch: 83 [23040/60000 (38%)]\tLoss: -2140.708203\n",
            "Train Epoch: 83 [24320/60000 (41%)]\tLoss: -2005.399414\n",
            "Train Epoch: 83 [25600/60000 (43%)]\tLoss: -1971.352344\n",
            "Train Epoch: 83 [26880/60000 (45%)]\tLoss: -1980.139844\n",
            "Train Epoch: 83 [28160/60000 (47%)]\tLoss: -2056.609570\n",
            "Train Epoch: 83 [29440/60000 (49%)]\tLoss: -1949.050586\n",
            "Train Epoch: 83 [30720/60000 (51%)]\tLoss: -1974.237109\n",
            "Train Epoch: 83 [32000/60000 (53%)]\tLoss: -2048.436914\n",
            "Train Epoch: 83 [33280/60000 (55%)]\tLoss: -1988.655273\n",
            "Train Epoch: 83 [34560/60000 (58%)]\tLoss: -2040.404297\n",
            "Train Epoch: 83 [35840/60000 (60%)]\tLoss: -1985.173047\n",
            "Train Epoch: 83 [37120/60000 (62%)]\tLoss: -2088.203906\n",
            "Train Epoch: 83 [38400/60000 (64%)]\tLoss: -1973.528125\n",
            "Train Epoch: 83 [39680/60000 (66%)]\tLoss: -1980.429297\n",
            "Train Epoch: 83 [40960/60000 (68%)]\tLoss: -2078.754883\n",
            "Train Epoch: 83 [42240/60000 (70%)]\tLoss: -2025.689648\n",
            "Train Epoch: 83 [43520/60000 (72%)]\tLoss: -2026.734961\n",
            "Train Epoch: 83 [44800/60000 (75%)]\tLoss: -2014.841797\n",
            "Train Epoch: 83 [46080/60000 (77%)]\tLoss: -2053.381445\n",
            "Train Epoch: 83 [47360/60000 (79%)]\tLoss: -2046.544336\n",
            "Train Epoch: 83 [48640/60000 (81%)]\tLoss: -2075.173438\n",
            "Train Epoch: 83 [49920/60000 (83%)]\tLoss: -2011.208594\n",
            "Train Epoch: 83 [51200/60000 (85%)]\tLoss: -2179.530664\n",
            "Train Epoch: 83 [52480/60000 (87%)]\tLoss: -1955.446680\n",
            "Train Epoch: 83 [53760/60000 (90%)]\tLoss: -2001.568945\n",
            "Train Epoch: 83 [55040/60000 (92%)]\tLoss: -1970.300781\n",
            "Train Epoch: 83 [56320/60000 (94%)]\tLoss: -1948.335938\n",
            "Train Epoch: 83 [57600/60000 (96%)]\tLoss: -1971.755859\n",
            "Train Epoch: 83 [58880/60000 (98%)]\tLoss: -1937.094922\n",
            "====> Epoch: 83 Average loss: -20286.6932\n",
            "Train Epoch: 84 [0/60000 (0%)]\tLoss: -19858.367188\n",
            "Train Epoch: 84 [1280/60000 (2%)]\tLoss: -2195.527344\n",
            "Train Epoch: 84 [2560/60000 (4%)]\tLoss: -1937.157422\n",
            "Train Epoch: 84 [3840/60000 (6%)]\tLoss: -1969.402148\n",
            "Train Epoch: 84 [5120/60000 (9%)]\tLoss: -2070.561523\n",
            "Train Epoch: 84 [6400/60000 (11%)]\tLoss: -2013.243359\n",
            "Train Epoch: 84 [7680/60000 (13%)]\tLoss: -2002.855469\n",
            "Train Epoch: 84 [8960/60000 (15%)]\tLoss: -2014.859766\n",
            "Train Epoch: 84 [10240/60000 (17%)]\tLoss: -2313.445898\n",
            "Train Epoch: 84 [11520/60000 (19%)]\tLoss: -2029.459961\n",
            "Train Epoch: 84 [12800/60000 (21%)]\tLoss: -2038.489258\n",
            "Train Epoch: 84 [14080/60000 (23%)]\tLoss: -2191.465234\n",
            "Train Epoch: 84 [15360/60000 (26%)]\tLoss: -1990.210352\n",
            "Train Epoch: 84 [16640/60000 (28%)]\tLoss: -1945.640430\n",
            "Train Epoch: 84 [17920/60000 (30%)]\tLoss: -1995.762109\n",
            "Train Epoch: 84 [19200/60000 (32%)]\tLoss: -2166.721680\n",
            "Train Epoch: 84 [20480/60000 (34%)]\tLoss: -1935.741016\n",
            "Train Epoch: 84 [21760/60000 (36%)]\tLoss: -2243.694336\n",
            "Train Epoch: 84 [23040/60000 (38%)]\tLoss: -2132.720313\n",
            "Train Epoch: 84 [24320/60000 (41%)]\tLoss: -2046.467187\n",
            "Train Epoch: 84 [25600/60000 (43%)]\tLoss: -1979.656641\n",
            "Train Epoch: 84 [26880/60000 (45%)]\tLoss: -1952.333789\n",
            "Train Epoch: 84 [28160/60000 (47%)]\tLoss: -2034.100000\n",
            "Train Epoch: 84 [29440/60000 (49%)]\tLoss: -1958.187305\n",
            "Train Epoch: 84 [30720/60000 (51%)]\tLoss: -1969.376953\n",
            "Train Epoch: 84 [32000/60000 (53%)]\tLoss: -2035.926172\n",
            "Train Epoch: 84 [33280/60000 (55%)]\tLoss: -1989.318359\n",
            "Train Epoch: 84 [34560/60000 (58%)]\tLoss: -2006.654102\n",
            "Train Epoch: 84 [35840/60000 (60%)]\tLoss: -2000.400586\n",
            "Train Epoch: 84 [37120/60000 (62%)]\tLoss: -2085.525000\n",
            "Train Epoch: 84 [38400/60000 (64%)]\tLoss: -1972.279687\n",
            "Train Epoch: 84 [39680/60000 (66%)]\tLoss: -1972.332813\n",
            "Train Epoch: 84 [40960/60000 (68%)]\tLoss: -2060.894141\n",
            "Train Epoch: 84 [42240/60000 (70%)]\tLoss: -2025.098633\n",
            "Train Epoch: 84 [43520/60000 (72%)]\tLoss: -1958.428320\n",
            "Train Epoch: 84 [44800/60000 (75%)]\tLoss: -2033.744141\n",
            "Train Epoch: 84 [46080/60000 (77%)]\tLoss: -2091.217578\n",
            "Train Epoch: 84 [47360/60000 (79%)]\tLoss: -2072.254687\n",
            "Train Epoch: 84 [48640/60000 (81%)]\tLoss: -2057.041406\n",
            "Train Epoch: 84 [49920/60000 (83%)]\tLoss: -2014.294336\n",
            "Train Epoch: 84 [51200/60000 (85%)]\tLoss: -2177.495508\n",
            "Train Epoch: 84 [52480/60000 (87%)]\tLoss: -1958.340430\n",
            "Train Epoch: 84 [53760/60000 (90%)]\tLoss: -1968.736133\n",
            "Train Epoch: 84 [55040/60000 (92%)]\tLoss: -1979.117188\n",
            "Train Epoch: 84 [56320/60000 (94%)]\tLoss: -1954.767969\n",
            "Train Epoch: 84 [57600/60000 (96%)]\tLoss: -1974.067578\n",
            "Train Epoch: 84 [58880/60000 (98%)]\tLoss: -1950.622656\n",
            "====> Epoch: 84 Average loss: -20275.4388\n",
            "Train Epoch: 85 [0/60000 (0%)]\tLoss: -19805.484375\n",
            "Train Epoch: 85 [1280/60000 (2%)]\tLoss: -2200.743555\n",
            "Train Epoch: 85 [2560/60000 (4%)]\tLoss: -1962.099023\n",
            "Train Epoch: 85 [3840/60000 (6%)]\tLoss: -1981.190430\n",
            "Train Epoch: 85 [5120/60000 (9%)]\tLoss: -2128.986328\n",
            "Train Epoch: 85 [6400/60000 (11%)]\tLoss: -2003.589844\n",
            "Train Epoch: 85 [7680/60000 (13%)]\tLoss: -1992.729492\n",
            "Train Epoch: 85 [8960/60000 (15%)]\tLoss: -2032.459375\n",
            "Train Epoch: 85 [10240/60000 (17%)]\tLoss: -2328.767773\n",
            "Train Epoch: 85 [11520/60000 (19%)]\tLoss: -1996.917578\n",
            "Train Epoch: 85 [12800/60000 (21%)]\tLoss: -2032.142969\n",
            "Train Epoch: 85 [14080/60000 (23%)]\tLoss: -2218.103320\n",
            "Train Epoch: 85 [15360/60000 (26%)]\tLoss: -1974.717773\n",
            "Train Epoch: 85 [16640/60000 (28%)]\tLoss: -1946.439258\n",
            "Train Epoch: 85 [17920/60000 (30%)]\tLoss: -2008.358203\n",
            "Train Epoch: 85 [19200/60000 (32%)]\tLoss: -2144.348633\n",
            "Train Epoch: 85 [20480/60000 (34%)]\tLoss: -1956.775000\n",
            "Train Epoch: 85 [21760/60000 (36%)]\tLoss: -2274.621680\n",
            "Train Epoch: 85 [23040/60000 (38%)]\tLoss: -2145.376367\n",
            "Train Epoch: 85 [24320/60000 (41%)]\tLoss: -2047.646289\n",
            "Train Epoch: 85 [25600/60000 (43%)]\tLoss: -1950.140625\n",
            "Train Epoch: 85 [26880/60000 (45%)]\tLoss: -1973.461328\n",
            "Train Epoch: 85 [28160/60000 (47%)]\tLoss: -2030.498047\n",
            "Train Epoch: 85 [29440/60000 (49%)]\tLoss: -1967.501563\n",
            "Train Epoch: 85 [30720/60000 (51%)]\tLoss: -1955.457227\n",
            "Train Epoch: 85 [32000/60000 (53%)]\tLoss: -2080.003906\n",
            "Train Epoch: 85 [33280/60000 (55%)]\tLoss: -1979.882812\n",
            "Train Epoch: 85 [34560/60000 (58%)]\tLoss: -2044.612891\n",
            "Train Epoch: 85 [35840/60000 (60%)]\tLoss: -2001.589258\n",
            "Train Epoch: 85 [37120/60000 (62%)]\tLoss: -2059.301758\n",
            "Train Epoch: 85 [38400/60000 (64%)]\tLoss: -1986.876172\n",
            "Train Epoch: 85 [39680/60000 (66%)]\tLoss: -1980.278906\n",
            "Train Epoch: 85 [40960/60000 (68%)]\tLoss: -2067.874023\n",
            "Train Epoch: 85 [42240/60000 (70%)]\tLoss: -2031.444922\n",
            "Train Epoch: 85 [43520/60000 (72%)]\tLoss: -1997.243945\n",
            "Train Epoch: 85 [44800/60000 (75%)]\tLoss: -2013.226953\n",
            "Train Epoch: 85 [46080/60000 (77%)]\tLoss: -2023.945117\n",
            "Train Epoch: 85 [47360/60000 (79%)]\tLoss: -2066.131055\n",
            "Train Epoch: 85 [48640/60000 (81%)]\tLoss: -2084.513672\n",
            "Train Epoch: 85 [49920/60000 (83%)]\tLoss: -2008.422266\n",
            "Train Epoch: 85 [51200/60000 (85%)]\tLoss: -2155.873047\n",
            "Train Epoch: 85 [52480/60000 (87%)]\tLoss: -1951.809766\n",
            "Train Epoch: 85 [53760/60000 (90%)]\tLoss: -1979.119141\n",
            "Train Epoch: 85 [55040/60000 (92%)]\tLoss: -1979.021484\n",
            "Train Epoch: 85 [56320/60000 (94%)]\tLoss: -1947.154492\n",
            "Train Epoch: 85 [57600/60000 (96%)]\tLoss: -1973.985352\n",
            "Train Epoch: 85 [58880/60000 (98%)]\tLoss: -1928.468555\n",
            "====> Epoch: 85 Average loss: -20283.6857\n",
            "Train Epoch: 86 [0/60000 (0%)]\tLoss: -20010.886719\n",
            "Train Epoch: 86 [1280/60000 (2%)]\tLoss: -2171.730273\n",
            "Train Epoch: 86 [2560/60000 (4%)]\tLoss: -1964.947461\n",
            "Train Epoch: 86 [3840/60000 (6%)]\tLoss: -1986.820508\n",
            "Train Epoch: 86 [5120/60000 (9%)]\tLoss: -2177.609180\n",
            "Train Epoch: 86 [6400/60000 (11%)]\tLoss: -1957.347070\n",
            "Train Epoch: 86 [7680/60000 (13%)]\tLoss: -1991.195898\n",
            "Train Epoch: 86 [8960/60000 (15%)]\tLoss: -2032.337109\n",
            "Train Epoch: 86 [10240/60000 (17%)]\tLoss: -2326.616406\n",
            "Train Epoch: 86 [11520/60000 (19%)]\tLoss: -1991.631250\n",
            "Train Epoch: 86 [12800/60000 (21%)]\tLoss: -2031.239453\n",
            "Train Epoch: 86 [14080/60000 (23%)]\tLoss: -2187.508594\n",
            "Train Epoch: 86 [15360/60000 (26%)]\tLoss: -1985.226953\n",
            "Train Epoch: 86 [16640/60000 (28%)]\tLoss: -1943.693359\n",
            "Train Epoch: 86 [17920/60000 (30%)]\tLoss: -1981.199023\n",
            "Train Epoch: 86 [19200/60000 (32%)]\tLoss: -2117.153320\n",
            "Train Epoch: 86 [20480/60000 (34%)]\tLoss: -1944.452734\n",
            "Train Epoch: 86 [21760/60000 (36%)]\tLoss: -2258.847461\n",
            "Train Epoch: 86 [23040/60000 (38%)]\tLoss: -2139.337695\n",
            "Train Epoch: 86 [24320/60000 (41%)]\tLoss: -2045.171094\n",
            "Train Epoch: 86 [25600/60000 (43%)]\tLoss: -1987.394336\n",
            "Train Epoch: 86 [26880/60000 (45%)]\tLoss: -1969.220508\n",
            "Train Epoch: 86 [28160/60000 (47%)]\tLoss: -2040.410547\n",
            "Train Epoch: 86 [29440/60000 (49%)]\tLoss: -1959.647656\n",
            "Train Epoch: 86 [30720/60000 (51%)]\tLoss: -1977.887305\n",
            "Train Epoch: 86 [32000/60000 (53%)]\tLoss: -2054.639062\n",
            "Train Epoch: 86 [33280/60000 (55%)]\tLoss: -1983.650781\n",
            "Train Epoch: 86 [34560/60000 (58%)]\tLoss: -2068.102930\n",
            "Train Epoch: 86 [35840/60000 (60%)]\tLoss: -2001.543555\n",
            "Train Epoch: 86 [37120/60000 (62%)]\tLoss: -2117.128320\n",
            "Train Epoch: 86 [38400/60000 (64%)]\tLoss: -1979.042187\n",
            "Train Epoch: 86 [39680/60000 (66%)]\tLoss: -1977.677734\n",
            "Train Epoch: 86 [40960/60000 (68%)]\tLoss: -2067.777930\n",
            "Train Epoch: 86 [42240/60000 (70%)]\tLoss: -2018.767188\n",
            "Train Epoch: 86 [43520/60000 (72%)]\tLoss: -1961.031641\n",
            "Train Epoch: 86 [44800/60000 (75%)]\tLoss: -2010.805664\n",
            "Train Epoch: 86 [46080/60000 (77%)]\tLoss: -2032.526172\n",
            "Train Epoch: 86 [47360/60000 (79%)]\tLoss: -2086.969141\n",
            "Train Epoch: 86 [48640/60000 (81%)]\tLoss: -2061.771680\n",
            "Train Epoch: 86 [49920/60000 (83%)]\tLoss: -2022.171094\n",
            "Train Epoch: 86 [51200/60000 (85%)]\tLoss: -2171.397461\n",
            "Train Epoch: 86 [52480/60000 (87%)]\tLoss: -1958.415234\n",
            "Train Epoch: 86 [53760/60000 (90%)]\tLoss: -1980.008203\n",
            "Train Epoch: 86 [55040/60000 (92%)]\tLoss: -1975.144141\n",
            "Train Epoch: 86 [56320/60000 (94%)]\tLoss: -1949.543945\n",
            "Train Epoch: 86 [57600/60000 (96%)]\tLoss: -1978.738867\n",
            "Train Epoch: 86 [58880/60000 (98%)]\tLoss: -1926.805078\n",
            "====> Epoch: 86 Average loss: -20278.9458\n",
            "Train Epoch: 87 [0/60000 (0%)]\tLoss: -19921.642578\n",
            "Train Epoch: 87 [1280/60000 (2%)]\tLoss: -2211.770508\n",
            "Train Epoch: 87 [2560/60000 (4%)]\tLoss: -1940.273633\n",
            "Train Epoch: 87 [3840/60000 (6%)]\tLoss: -1975.187305\n",
            "Train Epoch: 87 [5120/60000 (9%)]\tLoss: -2144.562695\n",
            "Train Epoch: 87 [6400/60000 (11%)]\tLoss: -1999.536523\n",
            "Train Epoch: 87 [7680/60000 (13%)]\tLoss: -2000.174609\n",
            "Train Epoch: 87 [8960/60000 (15%)]\tLoss: -2057.339063\n",
            "Train Epoch: 87 [10240/60000 (17%)]\tLoss: -2306.736133\n",
            "Train Epoch: 87 [11520/60000 (19%)]\tLoss: -2012.122070\n",
            "Train Epoch: 87 [12800/60000 (21%)]\tLoss: -2043.293164\n",
            "Train Epoch: 87 [14080/60000 (23%)]\tLoss: -2208.739258\n",
            "Train Epoch: 87 [15360/60000 (26%)]\tLoss: -2000.019531\n",
            "Train Epoch: 87 [16640/60000 (28%)]\tLoss: -1944.790039\n",
            "Train Epoch: 87 [17920/60000 (30%)]\tLoss: -1950.154687\n",
            "Train Epoch: 87 [19200/60000 (32%)]\tLoss: -2119.455469\n",
            "Train Epoch: 87 [20480/60000 (34%)]\tLoss: -1950.018164\n",
            "Train Epoch: 87 [21760/60000 (36%)]\tLoss: -2284.196875\n",
            "Train Epoch: 87 [23040/60000 (38%)]\tLoss: -2094.285156\n",
            "Train Epoch: 87 [24320/60000 (41%)]\tLoss: -2057.558984\n",
            "Train Epoch: 87 [25600/60000 (43%)]\tLoss: -1983.170703\n",
            "Train Epoch: 87 [26880/60000 (45%)]\tLoss: -1971.196484\n",
            "Train Epoch: 87 [28160/60000 (47%)]\tLoss: -2040.949414\n",
            "Train Epoch: 87 [29440/60000 (49%)]\tLoss: -1957.405664\n",
            "Train Epoch: 87 [30720/60000 (51%)]\tLoss: -2002.934570\n",
            "Train Epoch: 87 [32000/60000 (53%)]\tLoss: -2072.265430\n",
            "Train Epoch: 87 [33280/60000 (55%)]\tLoss: -1985.938281\n",
            "Train Epoch: 87 [34560/60000 (58%)]\tLoss: -2035.237109\n",
            "Train Epoch: 87 [35840/60000 (60%)]\tLoss: -2005.829492\n",
            "Train Epoch: 87 [37120/60000 (62%)]\tLoss: -2112.159180\n",
            "Train Epoch: 87 [38400/60000 (64%)]\tLoss: -1970.381836\n",
            "Train Epoch: 87 [39680/60000 (66%)]\tLoss: -1972.103711\n",
            "Train Epoch: 87 [40960/60000 (68%)]\tLoss: -2070.891797\n",
            "Train Epoch: 87 [42240/60000 (70%)]\tLoss: -2007.295898\n",
            "Train Epoch: 87 [43520/60000 (72%)]\tLoss: -2027.988867\n",
            "Train Epoch: 87 [44800/60000 (75%)]\tLoss: -2016.096484\n",
            "Train Epoch: 87 [46080/60000 (77%)]\tLoss: -2030.096289\n",
            "Train Epoch: 87 [47360/60000 (79%)]\tLoss: -2053.818945\n",
            "Train Epoch: 87 [48640/60000 (81%)]\tLoss: -2054.744531\n",
            "Train Epoch: 87 [49920/60000 (83%)]\tLoss: -2008.722461\n",
            "Train Epoch: 87 [51200/60000 (85%)]\tLoss: -2216.223828\n",
            "Train Epoch: 87 [52480/60000 (87%)]\tLoss: -1937.723828\n",
            "Train Epoch: 87 [53760/60000 (90%)]\tLoss: -1971.773438\n",
            "Train Epoch: 87 [55040/60000 (92%)]\tLoss: -1974.009375\n",
            "Train Epoch: 87 [56320/60000 (94%)]\tLoss: -1943.298047\n",
            "Train Epoch: 87 [57600/60000 (96%)]\tLoss: -1981.567383\n",
            "Train Epoch: 87 [58880/60000 (98%)]\tLoss: -1931.145117\n",
            "====> Epoch: 87 Average loss: -20300.1003\n",
            "Train Epoch: 88 [0/60000 (0%)]\tLoss: -19820.400391\n",
            "Train Epoch: 88 [1280/60000 (2%)]\tLoss: -2192.163477\n",
            "Train Epoch: 88 [2560/60000 (4%)]\tLoss: -1951.102344\n",
            "Train Epoch: 88 [3840/60000 (6%)]\tLoss: -1981.182031\n",
            "Train Epoch: 88 [5120/60000 (9%)]\tLoss: -2050.253516\n",
            "Train Epoch: 88 [6400/60000 (11%)]\tLoss: -2013.444336\n",
            "Train Epoch: 88 [7680/60000 (13%)]\tLoss: -1992.358594\n",
            "Train Epoch: 88 [8960/60000 (15%)]\tLoss: -2052.079492\n",
            "Train Epoch: 88 [10240/60000 (17%)]\tLoss: -2351.469922\n",
            "Train Epoch: 88 [11520/60000 (19%)]\tLoss: -1974.352930\n",
            "Train Epoch: 88 [12800/60000 (21%)]\tLoss: -2046.294141\n",
            "Train Epoch: 88 [14080/60000 (23%)]\tLoss: -2211.841797\n",
            "Train Epoch: 88 [15360/60000 (26%)]\tLoss: -1993.183984\n",
            "Train Epoch: 88 [16640/60000 (28%)]\tLoss: -1944.006445\n",
            "Train Epoch: 88 [17920/60000 (30%)]\tLoss: -1965.548438\n",
            "Train Epoch: 88 [19200/60000 (32%)]\tLoss: -2143.168555\n",
            "Train Epoch: 88 [20480/60000 (34%)]\tLoss: -1953.190820\n",
            "Train Epoch: 88 [21760/60000 (36%)]\tLoss: -2266.866992\n",
            "Train Epoch: 88 [23040/60000 (38%)]\tLoss: -2150.020898\n",
            "Train Epoch: 88 [24320/60000 (41%)]\tLoss: -2063.529102\n",
            "Train Epoch: 88 [25600/60000 (43%)]\tLoss: -1974.054297\n",
            "Train Epoch: 88 [26880/60000 (45%)]\tLoss: -1976.968359\n",
            "Train Epoch: 88 [28160/60000 (47%)]\tLoss: -2080.218945\n",
            "Train Epoch: 88 [29440/60000 (49%)]\tLoss: -1960.769727\n",
            "Train Epoch: 88 [30720/60000 (51%)]\tLoss: -1971.912109\n",
            "Train Epoch: 88 [32000/60000 (53%)]\tLoss: -2036.152148\n",
            "Train Epoch: 88 [33280/60000 (55%)]\tLoss: -1962.651562\n",
            "Train Epoch: 88 [34560/60000 (58%)]\tLoss: -2023.896289\n",
            "Train Epoch: 88 [35840/60000 (60%)]\tLoss: -2002.721094\n",
            "Train Epoch: 88 [37120/60000 (62%)]\tLoss: -2066.764453\n",
            "Train Epoch: 88 [38400/60000 (64%)]\tLoss: -1986.282422\n",
            "Train Epoch: 88 [39680/60000 (66%)]\tLoss: -1972.440234\n",
            "Train Epoch: 88 [40960/60000 (68%)]\tLoss: -2092.321680\n",
            "Train Epoch: 88 [42240/60000 (70%)]\tLoss: -2071.746094\n",
            "Train Epoch: 88 [43520/60000 (72%)]\tLoss: -1992.239063\n",
            "Train Epoch: 88 [44800/60000 (75%)]\tLoss: -2027.016406\n",
            "Train Epoch: 88 [46080/60000 (77%)]\tLoss: -2032.944141\n",
            "Train Epoch: 88 [47360/60000 (79%)]\tLoss: -2037.516016\n",
            "Train Epoch: 88 [48640/60000 (81%)]\tLoss: -2086.533984\n",
            "Train Epoch: 88 [49920/60000 (83%)]\tLoss: -1995.964844\n",
            "Train Epoch: 88 [51200/60000 (85%)]\tLoss: -2186.228711\n",
            "Train Epoch: 88 [52480/60000 (87%)]\tLoss: -1951.770898\n",
            "Train Epoch: 88 [53760/60000 (90%)]\tLoss: -1986.916211\n",
            "Train Epoch: 88 [55040/60000 (92%)]\tLoss: -1986.261914\n",
            "Train Epoch: 88 [56320/60000 (94%)]\tLoss: -1943.825977\n",
            "Train Epoch: 88 [57600/60000 (96%)]\tLoss: -1986.240039\n",
            "Train Epoch: 88 [58880/60000 (98%)]\tLoss: -1945.612695\n",
            "====> Epoch: 88 Average loss: -20280.9667\n",
            "Train Epoch: 89 [0/60000 (0%)]\tLoss: -19809.361328\n",
            "Train Epoch: 89 [1280/60000 (2%)]\tLoss: -2146.438477\n",
            "Train Epoch: 89 [2560/60000 (4%)]\tLoss: -1936.511523\n",
            "Train Epoch: 89 [3840/60000 (6%)]\tLoss: -1989.410547\n",
            "Train Epoch: 89 [5120/60000 (9%)]\tLoss: -2047.642383\n",
            "Train Epoch: 89 [6400/60000 (11%)]\tLoss: -2033.961133\n",
            "Train Epoch: 89 [7680/60000 (13%)]\tLoss: -2027.960547\n",
            "Train Epoch: 89 [8960/60000 (15%)]\tLoss: -2089.101172\n",
            "Train Epoch: 89 [10240/60000 (17%)]\tLoss: -2329.602344\n",
            "Train Epoch: 89 [11520/60000 (19%)]\tLoss: -2003.561133\n",
            "Train Epoch: 89 [12800/60000 (21%)]\tLoss: -2043.311133\n",
            "Train Epoch: 89 [14080/60000 (23%)]\tLoss: -2220.594531\n",
            "Train Epoch: 89 [15360/60000 (26%)]\tLoss: -1986.714453\n",
            "Train Epoch: 89 [16640/60000 (28%)]\tLoss: -1949.960547\n",
            "Train Epoch: 89 [17920/60000 (30%)]\tLoss: -1980.305273\n",
            "Train Epoch: 89 [19200/60000 (32%)]\tLoss: -2140.628320\n",
            "Train Epoch: 89 [20480/60000 (34%)]\tLoss: -1959.825586\n",
            "Train Epoch: 89 [21760/60000 (36%)]\tLoss: -2258.922461\n",
            "Train Epoch: 89 [23040/60000 (38%)]\tLoss: -2139.618164\n",
            "Train Epoch: 89 [24320/60000 (41%)]\tLoss: -2052.858789\n",
            "Train Epoch: 89 [25600/60000 (43%)]\tLoss: -1977.674805\n",
            "Train Epoch: 89 [26880/60000 (45%)]\tLoss: -1962.336914\n",
            "Train Epoch: 89 [28160/60000 (47%)]\tLoss: -2049.912891\n",
            "Train Epoch: 89 [29440/60000 (49%)]\tLoss: -1957.821680\n",
            "Train Epoch: 89 [30720/60000 (51%)]\tLoss: -1976.860156\n",
            "Train Epoch: 89 [32000/60000 (53%)]\tLoss: -2043.195117\n",
            "Train Epoch: 89 [33280/60000 (55%)]\tLoss: -1988.877734\n",
            "Train Epoch: 89 [34560/60000 (58%)]\tLoss: -1991.345898\n",
            "Train Epoch: 89 [35840/60000 (60%)]\tLoss: -1988.850000\n",
            "Train Epoch: 89 [37120/60000 (62%)]\tLoss: -2059.257227\n",
            "Train Epoch: 89 [38400/60000 (64%)]\tLoss: -1974.369531\n",
            "Train Epoch: 89 [39680/60000 (66%)]\tLoss: -1975.643945\n",
            "Train Epoch: 89 [40960/60000 (68%)]\tLoss: -2052.863477\n",
            "Train Epoch: 89 [42240/60000 (70%)]\tLoss: -2041.230664\n",
            "Train Epoch: 89 [43520/60000 (72%)]\tLoss: -1988.775195\n",
            "Train Epoch: 89 [44800/60000 (75%)]\tLoss: -2031.776562\n",
            "Train Epoch: 89 [46080/60000 (77%)]\tLoss: -2065.372656\n",
            "Train Epoch: 89 [47360/60000 (79%)]\tLoss: -2077.483984\n",
            "Train Epoch: 89 [48640/60000 (81%)]\tLoss: -2046.104687\n",
            "Train Epoch: 89 [49920/60000 (83%)]\tLoss: -2003.546484\n",
            "Train Epoch: 89 [51200/60000 (85%)]\tLoss: -2206.126563\n",
            "Train Epoch: 89 [52480/60000 (87%)]\tLoss: -1936.469141\n",
            "Train Epoch: 89 [53760/60000 (90%)]\tLoss: -1995.546875\n",
            "Train Epoch: 89 [55040/60000 (92%)]\tLoss: -1991.933789\n",
            "Train Epoch: 89 [56320/60000 (94%)]\tLoss: -1943.678711\n",
            "Train Epoch: 89 [57600/60000 (96%)]\tLoss: -1990.816992\n",
            "Train Epoch: 89 [58880/60000 (98%)]\tLoss: -1958.866211\n",
            "====> Epoch: 89 Average loss: -20267.6822\n",
            "Train Epoch: 90 [0/60000 (0%)]\tLoss: -19940.949219\n",
            "Train Epoch: 90 [1280/60000 (2%)]\tLoss: -2217.371094\n",
            "Train Epoch: 90 [2560/60000 (4%)]\tLoss: -1945.452539\n",
            "Train Epoch: 90 [3840/60000 (6%)]\tLoss: -2000.532422\n",
            "Train Epoch: 90 [5120/60000 (9%)]\tLoss: -2077.532227\n",
            "Train Epoch: 90 [6400/60000 (11%)]\tLoss: -2032.662500\n",
            "Train Epoch: 90 [7680/60000 (13%)]\tLoss: -2013.599805\n",
            "Train Epoch: 90 [8960/60000 (15%)]\tLoss: -2053.414844\n",
            "Train Epoch: 90 [10240/60000 (17%)]\tLoss: -2332.199414\n",
            "Train Epoch: 90 [11520/60000 (19%)]\tLoss: -2029.070508\n",
            "Train Epoch: 90 [12800/60000 (21%)]\tLoss: -2002.410352\n",
            "Train Epoch: 90 [14080/60000 (23%)]\tLoss: -2200.733984\n",
            "Train Epoch: 90 [15360/60000 (26%)]\tLoss: -1973.757227\n",
            "Train Epoch: 90 [16640/60000 (28%)]\tLoss: -1947.265820\n",
            "Train Epoch: 90 [17920/60000 (30%)]\tLoss: -1958.501367\n",
            "Train Epoch: 90 [19200/60000 (32%)]\tLoss: -2141.338281\n",
            "Train Epoch: 90 [20480/60000 (34%)]\tLoss: -1981.470117\n",
            "Train Epoch: 90 [21760/60000 (36%)]\tLoss: -2268.732812\n",
            "Train Epoch: 90 [23040/60000 (38%)]\tLoss: -2157.906250\n",
            "Train Epoch: 90 [24320/60000 (41%)]\tLoss: -2044.630469\n",
            "Train Epoch: 90 [25600/60000 (43%)]\tLoss: -1959.719336\n",
            "Train Epoch: 90 [26880/60000 (45%)]\tLoss: -1960.793555\n",
            "Train Epoch: 90 [28160/60000 (47%)]\tLoss: -2015.533789\n",
            "Train Epoch: 90 [29440/60000 (49%)]\tLoss: -1964.671875\n",
            "Train Epoch: 90 [30720/60000 (51%)]\tLoss: -1987.493359\n",
            "Train Epoch: 90 [32000/60000 (53%)]\tLoss: -2066.021289\n",
            "Train Epoch: 90 [33280/60000 (55%)]\tLoss: -1984.221680\n",
            "Train Epoch: 90 [34560/60000 (58%)]\tLoss: -2013.720117\n",
            "Train Epoch: 90 [35840/60000 (60%)]\tLoss: -1992.663672\n",
            "Train Epoch: 90 [37120/60000 (62%)]\tLoss: -2028.331445\n",
            "Train Epoch: 90 [38400/60000 (64%)]\tLoss: -1989.782031\n",
            "Train Epoch: 90 [39680/60000 (66%)]\tLoss: -1988.104883\n",
            "Train Epoch: 90 [40960/60000 (68%)]\tLoss: -2064.035547\n",
            "Train Epoch: 90 [42240/60000 (70%)]\tLoss: -2039.172266\n",
            "Train Epoch: 90 [43520/60000 (72%)]\tLoss: -2026.758984\n",
            "Train Epoch: 90 [44800/60000 (75%)]\tLoss: -2016.144141\n",
            "Train Epoch: 90 [46080/60000 (77%)]\tLoss: -2109.056055\n",
            "Train Epoch: 90 [47360/60000 (79%)]\tLoss: -2100.462891\n",
            "Train Epoch: 90 [48640/60000 (81%)]\tLoss: -2042.812305\n",
            "Train Epoch: 90 [49920/60000 (83%)]\tLoss: -2012.961328\n",
            "Train Epoch: 90 [51200/60000 (85%)]\tLoss: -2179.681055\n",
            "Train Epoch: 90 [52480/60000 (87%)]\tLoss: -1932.131445\n",
            "Train Epoch: 90 [53760/60000 (90%)]\tLoss: -1985.982422\n",
            "Train Epoch: 90 [55040/60000 (92%)]\tLoss: -1972.668945\n",
            "Train Epoch: 90 [56320/60000 (94%)]\tLoss: -1949.625781\n",
            "Train Epoch: 90 [57600/60000 (96%)]\tLoss: -1976.825000\n",
            "Train Epoch: 90 [58880/60000 (98%)]\tLoss: -1951.712305\n",
            "====> Epoch: 90 Average loss: -20291.2643\n",
            "Train Epoch: 91 [0/60000 (0%)]\tLoss: -19829.396484\n",
            "Train Epoch: 91 [1280/60000 (2%)]\tLoss: -2202.506836\n",
            "Train Epoch: 91 [2560/60000 (4%)]\tLoss: -1936.632031\n",
            "Train Epoch: 91 [3840/60000 (6%)]\tLoss: -1996.332813\n",
            "Train Epoch: 91 [5120/60000 (9%)]\tLoss: -2048.925977\n",
            "Train Epoch: 91 [6400/60000 (11%)]\tLoss: -2004.151367\n",
            "Train Epoch: 91 [7680/60000 (13%)]\tLoss: -2027.151367\n",
            "Train Epoch: 91 [8960/60000 (15%)]\tLoss: -2103.720117\n",
            "Train Epoch: 91 [10240/60000 (17%)]\tLoss: -2387.852930\n",
            "Train Epoch: 91 [11520/60000 (19%)]\tLoss: -1982.952930\n",
            "Train Epoch: 91 [12800/60000 (21%)]\tLoss: -2039.156836\n",
            "Train Epoch: 91 [14080/60000 (23%)]\tLoss: -2193.764844\n",
            "Train Epoch: 91 [15360/60000 (26%)]\tLoss: -1990.957031\n",
            "Train Epoch: 91 [16640/60000 (28%)]\tLoss: -1945.127734\n",
            "Train Epoch: 91 [17920/60000 (30%)]\tLoss: -1992.533984\n",
            "Train Epoch: 91 [19200/60000 (32%)]\tLoss: -2169.192969\n",
            "Train Epoch: 91 [20480/60000 (34%)]\tLoss: -1964.681641\n",
            "Train Epoch: 91 [21760/60000 (36%)]\tLoss: -2292.341406\n",
            "Train Epoch: 91 [23040/60000 (38%)]\tLoss: -2134.594141\n",
            "Train Epoch: 91 [24320/60000 (41%)]\tLoss: -2041.058203\n",
            "Train Epoch: 91 [25600/60000 (43%)]\tLoss: -1941.395703\n",
            "Train Epoch: 91 [26880/60000 (45%)]\tLoss: -1964.375195\n",
            "Train Epoch: 91 [28160/60000 (47%)]\tLoss: -2083.407617\n",
            "Train Epoch: 91 [29440/60000 (49%)]\tLoss: -1956.301367\n",
            "Train Epoch: 91 [30720/60000 (51%)]\tLoss: -1993.243945\n",
            "Train Epoch: 91 [32000/60000 (53%)]\tLoss: -2029.874414\n",
            "Train Epoch: 91 [33280/60000 (55%)]\tLoss: -1984.628906\n",
            "Train Epoch: 91 [34560/60000 (58%)]\tLoss: -2029.716992\n",
            "Train Epoch: 91 [35840/60000 (60%)]\tLoss: -1988.201758\n",
            "Train Epoch: 91 [37120/60000 (62%)]\tLoss: -2061.558789\n",
            "Train Epoch: 91 [38400/60000 (64%)]\tLoss: -2005.868164\n",
            "Train Epoch: 91 [39680/60000 (66%)]\tLoss: -1970.546875\n",
            "Train Epoch: 91 [40960/60000 (68%)]\tLoss: -2063.000195\n",
            "Train Epoch: 91 [42240/60000 (70%)]\tLoss: -2050.611133\n",
            "Train Epoch: 91 [43520/60000 (72%)]\tLoss: -2061.874219\n",
            "Train Epoch: 91 [44800/60000 (75%)]\tLoss: -2016.809375\n",
            "Train Epoch: 91 [46080/60000 (77%)]\tLoss: -2103.585938\n",
            "Train Epoch: 91 [47360/60000 (79%)]\tLoss: -2126.796289\n",
            "Train Epoch: 91 [48640/60000 (81%)]\tLoss: -2078.729492\n",
            "Train Epoch: 91 [49920/60000 (83%)]\tLoss: -2007.956250\n",
            "Train Epoch: 91 [51200/60000 (85%)]\tLoss: -2206.964453\n",
            "Train Epoch: 91 [52480/60000 (87%)]\tLoss: -1929.605859\n",
            "Train Epoch: 91 [53760/60000 (90%)]\tLoss: -1981.954492\n",
            "Train Epoch: 91 [55040/60000 (92%)]\tLoss: -1970.414062\n",
            "Train Epoch: 91 [56320/60000 (94%)]\tLoss: -1948.482617\n",
            "Train Epoch: 91 [57600/60000 (96%)]\tLoss: -1966.902148\n",
            "Train Epoch: 91 [58880/60000 (98%)]\tLoss: -1948.149805\n",
            "====> Epoch: 91 Average loss: -20333.2718\n",
            "Train Epoch: 92 [0/60000 (0%)]\tLoss: -19710.347656\n",
            "Train Epoch: 92 [1280/60000 (2%)]\tLoss: -2143.333203\n",
            "Train Epoch: 92 [2560/60000 (4%)]\tLoss: -1978.552930\n",
            "Train Epoch: 92 [3840/60000 (6%)]\tLoss: -2000.860352\n",
            "Train Epoch: 92 [5120/60000 (9%)]\tLoss: -2077.264844\n",
            "Train Epoch: 92 [6400/60000 (11%)]\tLoss: -1979.051758\n",
            "Train Epoch: 92 [7680/60000 (13%)]\tLoss: -2002.804688\n",
            "Train Epoch: 92 [8960/60000 (15%)]\tLoss: -2080.825000\n",
            "Train Epoch: 92 [10240/60000 (17%)]\tLoss: -2353.966406\n",
            "Train Epoch: 92 [11520/60000 (19%)]\tLoss: -1993.984766\n",
            "Train Epoch: 92 [12800/60000 (21%)]\tLoss: -2025.230664\n",
            "Train Epoch: 92 [14080/60000 (23%)]\tLoss: -2175.995898\n",
            "Train Epoch: 92 [15360/60000 (26%)]\tLoss: -2002.730078\n",
            "Train Epoch: 92 [16640/60000 (28%)]\tLoss: -1948.167969\n",
            "Train Epoch: 92 [17920/60000 (30%)]\tLoss: -1997.694727\n",
            "Train Epoch: 92 [19200/60000 (32%)]\tLoss: -2170.473242\n",
            "Train Epoch: 92 [20480/60000 (34%)]\tLoss: -1961.888672\n",
            "Train Epoch: 92 [21760/60000 (36%)]\tLoss: -2267.449414\n",
            "Train Epoch: 92 [23040/60000 (38%)]\tLoss: -2140.238672\n",
            "Train Epoch: 92 [24320/60000 (41%)]\tLoss: -2040.612305\n",
            "Train Epoch: 92 [25600/60000 (43%)]\tLoss: -1947.091992\n",
            "Train Epoch: 92 [26880/60000 (45%)]\tLoss: -1964.025195\n",
            "Train Epoch: 92 [28160/60000 (47%)]\tLoss: -2052.020117\n",
            "Train Epoch: 92 [29440/60000 (49%)]\tLoss: -1982.973828\n",
            "Train Epoch: 92 [30720/60000 (51%)]\tLoss: -1991.533008\n",
            "Train Epoch: 92 [32000/60000 (53%)]\tLoss: -2087.183984\n",
            "Train Epoch: 92 [33280/60000 (55%)]\tLoss: -2004.518359\n",
            "Train Epoch: 92 [34560/60000 (58%)]\tLoss: -2037.632422\n",
            "Train Epoch: 92 [35840/60000 (60%)]\tLoss: -1995.507422\n",
            "Train Epoch: 92 [37120/60000 (62%)]\tLoss: -2083.798828\n",
            "Train Epoch: 92 [38400/60000 (64%)]\tLoss: -1997.245898\n",
            "Train Epoch: 92 [39680/60000 (66%)]\tLoss: -1984.217773\n",
            "Train Epoch: 92 [40960/60000 (68%)]\tLoss: -2057.519141\n",
            "Train Epoch: 92 [42240/60000 (70%)]\tLoss: -2059.403906\n",
            "Train Epoch: 92 [43520/60000 (72%)]\tLoss: -2030.718945\n",
            "Train Epoch: 92 [44800/60000 (75%)]\tLoss: -2024.889453\n",
            "Train Epoch: 92 [46080/60000 (77%)]\tLoss: -2103.437109\n",
            "Train Epoch: 92 [47360/60000 (79%)]\tLoss: -2110.874414\n",
            "Train Epoch: 92 [48640/60000 (81%)]\tLoss: -2083.303125\n",
            "Train Epoch: 92 [49920/60000 (83%)]\tLoss: -2017.805273\n",
            "Train Epoch: 92 [51200/60000 (85%)]\tLoss: -2238.712891\n",
            "Train Epoch: 92 [52480/60000 (87%)]\tLoss: -1929.969336\n",
            "Train Epoch: 92 [53760/60000 (90%)]\tLoss: -1996.220313\n",
            "Train Epoch: 92 [55040/60000 (92%)]\tLoss: -1971.331641\n",
            "Train Epoch: 92 [56320/60000 (94%)]\tLoss: -1945.764258\n",
            "Train Epoch: 92 [57600/60000 (96%)]\tLoss: -1970.874805\n",
            "Train Epoch: 92 [58880/60000 (98%)]\tLoss: -1975.360156\n",
            "====> Epoch: 92 Average loss: -20336.0682\n",
            "Train Epoch: 93 [0/60000 (0%)]\tLoss: -20179.251953\n",
            "Train Epoch: 93 [1280/60000 (2%)]\tLoss: -2170.671094\n",
            "Train Epoch: 93 [2560/60000 (4%)]\tLoss: -1981.358594\n",
            "Train Epoch: 93 [3840/60000 (6%)]\tLoss: -2009.448828\n",
            "Train Epoch: 93 [5120/60000 (9%)]\tLoss: -2073.808789\n",
            "Train Epoch: 93 [6400/60000 (11%)]\tLoss: -1979.506641\n",
            "Train Epoch: 93 [7680/60000 (13%)]\tLoss: -2021.007031\n",
            "Train Epoch: 93 [8960/60000 (15%)]\tLoss: -2100.788477\n",
            "Train Epoch: 93 [10240/60000 (17%)]\tLoss: -2412.280273\n",
            "Train Epoch: 93 [11520/60000 (19%)]\tLoss: -2053.866602\n",
            "Train Epoch: 93 [12800/60000 (21%)]\tLoss: -2029.832813\n",
            "Train Epoch: 93 [14080/60000 (23%)]\tLoss: -2249.472852\n",
            "Train Epoch: 93 [15360/60000 (26%)]\tLoss: -1973.834180\n",
            "Train Epoch: 93 [16640/60000 (28%)]\tLoss: -1943.843359\n",
            "Train Epoch: 93 [17920/60000 (30%)]\tLoss: -2022.316406\n",
            "Train Epoch: 93 [19200/60000 (32%)]\tLoss: -2130.454492\n",
            "Train Epoch: 93 [20480/60000 (34%)]\tLoss: -1975.807422\n",
            "Train Epoch: 93 [21760/60000 (36%)]\tLoss: -2238.813672\n",
            "Train Epoch: 93 [23040/60000 (38%)]\tLoss: -2135.720508\n",
            "Train Epoch: 93 [24320/60000 (41%)]\tLoss: -2055.743164\n",
            "Train Epoch: 93 [25600/60000 (43%)]\tLoss: -1932.567969\n",
            "Train Epoch: 93 [26880/60000 (45%)]\tLoss: -1986.148047\n",
            "Train Epoch: 93 [28160/60000 (47%)]\tLoss: -2043.306250\n",
            "Train Epoch: 93 [29440/60000 (49%)]\tLoss: -1984.166602\n",
            "Train Epoch: 93 [30720/60000 (51%)]\tLoss: -1978.043555\n",
            "Train Epoch: 93 [32000/60000 (53%)]\tLoss: -2114.503125\n",
            "Train Epoch: 93 [33280/60000 (55%)]\tLoss: -2017.876172\n",
            "Train Epoch: 93 [34560/60000 (58%)]\tLoss: -2000.496680\n",
            "Train Epoch: 93 [35840/60000 (60%)]\tLoss: -1990.438281\n",
            "Train Epoch: 93 [37120/60000 (62%)]\tLoss: -2064.354688\n",
            "Train Epoch: 93 [38400/60000 (64%)]\tLoss: -1976.960156\n",
            "Train Epoch: 93 [39680/60000 (66%)]\tLoss: -2008.420508\n",
            "Train Epoch: 93 [40960/60000 (68%)]\tLoss: -2063.341797\n",
            "Train Epoch: 93 [42240/60000 (70%)]\tLoss: -2082.295508\n",
            "Train Epoch: 93 [43520/60000 (72%)]\tLoss: -2043.018945\n",
            "Train Epoch: 93 [44800/60000 (75%)]\tLoss: -2035.712305\n",
            "Train Epoch: 93 [46080/60000 (77%)]\tLoss: -2114.806836\n",
            "Train Epoch: 93 [47360/60000 (79%)]\tLoss: -2106.601172\n",
            "Train Epoch: 93 [48640/60000 (81%)]\tLoss: -2110.944531\n",
            "Train Epoch: 93 [49920/60000 (83%)]\tLoss: -2012.772656\n",
            "Train Epoch: 93 [51200/60000 (85%)]\tLoss: -2234.270117\n",
            "Train Epoch: 93 [52480/60000 (87%)]\tLoss: -1951.077734\n",
            "Train Epoch: 93 [53760/60000 (90%)]\tLoss: -1979.112891\n",
            "Train Epoch: 93 [55040/60000 (92%)]\tLoss: -1978.511523\n",
            "Train Epoch: 93 [56320/60000 (94%)]\tLoss: -1955.496289\n",
            "Train Epoch: 93 [57600/60000 (96%)]\tLoss: -1985.250391\n",
            "Train Epoch: 93 [58880/60000 (98%)]\tLoss: -2030.188867\n",
            "====> Epoch: 93 Average loss: -20428.9728\n",
            "Train Epoch: 94 [0/60000 (0%)]\tLoss: -20482.896484\n",
            "Train Epoch: 94 [1280/60000 (2%)]\tLoss: -2137.916406\n",
            "Train Epoch: 94 [2560/60000 (4%)]\tLoss: -1958.048242\n",
            "Train Epoch: 94 [3840/60000 (6%)]\tLoss: -2026.280664\n",
            "Train Epoch: 94 [5120/60000 (9%)]\tLoss: -2215.010938\n",
            "Train Epoch: 94 [6400/60000 (11%)]\tLoss: -1989.167187\n",
            "Train Epoch: 94 [7680/60000 (13%)]\tLoss: -2054.870508\n",
            "Train Epoch: 94 [8960/60000 (15%)]\tLoss: -2055.983398\n",
            "Train Epoch: 94 [10240/60000 (17%)]\tLoss: -2392.742773\n",
            "Train Epoch: 94 [11520/60000 (19%)]\tLoss: -2020.847656\n",
            "Train Epoch: 94 [12800/60000 (21%)]\tLoss: -2048.741406\n",
            "Train Epoch: 94 [14080/60000 (23%)]\tLoss: -2361.811133\n",
            "Train Epoch: 94 [15360/60000 (26%)]\tLoss: -1992.019922\n",
            "Train Epoch: 94 [16640/60000 (28%)]\tLoss: -1953.717578\n",
            "Train Epoch: 94 [17920/60000 (30%)]\tLoss: -2042.587891\n",
            "Train Epoch: 94 [19200/60000 (32%)]\tLoss: -2149.374414\n",
            "Train Epoch: 94 [20480/60000 (34%)]\tLoss: -1989.726758\n",
            "Train Epoch: 94 [21760/60000 (36%)]\tLoss: -2236.157617\n",
            "Train Epoch: 94 [23040/60000 (38%)]\tLoss: -2117.435156\n",
            "Train Epoch: 94 [24320/60000 (41%)]\tLoss: -2068.185937\n",
            "Train Epoch: 94 [25600/60000 (43%)]\tLoss: -1965.339258\n",
            "Train Epoch: 94 [26880/60000 (45%)]\tLoss: -1997.214844\n",
            "Train Epoch: 94 [28160/60000 (47%)]\tLoss: -2104.886523\n",
            "Train Epoch: 94 [29440/60000 (49%)]\tLoss: -1963.870508\n",
            "Train Epoch: 94 [30720/60000 (51%)]\tLoss: -1992.824609\n",
            "Train Epoch: 94 [32000/60000 (53%)]\tLoss: -2085.556836\n",
            "Train Epoch: 94 [33280/60000 (55%)]\tLoss: -2027.019531\n",
            "Train Epoch: 94 [34560/60000 (58%)]\tLoss: -1992.129297\n",
            "Train Epoch: 94 [35840/60000 (60%)]\tLoss: -2002.703320\n",
            "Train Epoch: 94 [37120/60000 (62%)]\tLoss: -2122.602148\n",
            "Train Epoch: 94 [38400/60000 (64%)]\tLoss: -2014.290234\n",
            "Train Epoch: 94 [39680/60000 (66%)]\tLoss: -2059.177930\n",
            "Train Epoch: 94 [40960/60000 (68%)]\tLoss: -2054.229297\n",
            "Train Epoch: 94 [42240/60000 (70%)]\tLoss: -2046.590820\n",
            "Train Epoch: 94 [43520/60000 (72%)]\tLoss: -2040.253711\n",
            "Train Epoch: 94 [44800/60000 (75%)]\tLoss: -2067.630273\n",
            "Train Epoch: 94 [46080/60000 (77%)]\tLoss: -2112.042773\n",
            "Train Epoch: 94 [47360/60000 (79%)]\tLoss: -2158.250781\n",
            "Train Epoch: 94 [48640/60000 (81%)]\tLoss: -2124.079102\n",
            "Train Epoch: 94 [49920/60000 (83%)]\tLoss: -2006.395703\n",
            "Train Epoch: 94 [51200/60000 (85%)]\tLoss: -2236.067187\n",
            "Train Epoch: 94 [52480/60000 (87%)]\tLoss: -1935.380859\n",
            "Train Epoch: 94 [53760/60000 (90%)]\tLoss: -1990.968945\n",
            "Train Epoch: 94 [55040/60000 (92%)]\tLoss: -2017.375391\n",
            "Train Epoch: 94 [56320/60000 (94%)]\tLoss: -1966.588867\n",
            "Train Epoch: 94 [57600/60000 (96%)]\tLoss: -2014.298438\n",
            "Train Epoch: 94 [58880/60000 (98%)]\tLoss: -1996.980664\n",
            "====> Epoch: 94 Average loss: -20537.6268\n",
            "Train Epoch: 95 [0/60000 (0%)]\tLoss: -20270.089844\n",
            "Train Epoch: 95 [1280/60000 (2%)]\tLoss: -2172.521094\n",
            "Train Epoch: 95 [2560/60000 (4%)]\tLoss: -1978.967383\n",
            "Train Epoch: 95 [3840/60000 (6%)]\tLoss: -2032.271680\n",
            "Train Epoch: 95 [5120/60000 (9%)]\tLoss: -2230.635742\n",
            "Train Epoch: 95 [6400/60000 (11%)]\tLoss: -2000.388086\n",
            "Train Epoch: 95 [7680/60000 (13%)]\tLoss: -1996.338086\n",
            "Train Epoch: 95 [8960/60000 (15%)]\tLoss: -2090.013672\n",
            "Train Epoch: 95 [10240/60000 (17%)]\tLoss: -2366.125391\n",
            "Train Epoch: 95 [11520/60000 (19%)]\tLoss: -2040.462695\n",
            "Train Epoch: 95 [12800/60000 (21%)]\tLoss: -2057.561523\n",
            "Train Epoch: 95 [14080/60000 (23%)]\tLoss: -2378.941992\n",
            "Train Epoch: 95 [15360/60000 (26%)]\tLoss: -2019.108594\n",
            "Train Epoch: 95 [16640/60000 (28%)]\tLoss: -1939.503516\n",
            "Train Epoch: 95 [17920/60000 (30%)]\tLoss: -1982.897266\n",
            "Train Epoch: 95 [19200/60000 (32%)]\tLoss: -2137.641602\n",
            "Train Epoch: 95 [20480/60000 (34%)]\tLoss: -1996.457031\n",
            "Train Epoch: 95 [21760/60000 (36%)]\tLoss: -2300.513086\n",
            "Train Epoch: 95 [23040/60000 (38%)]\tLoss: -2152.476367\n",
            "Train Epoch: 95 [24320/60000 (41%)]\tLoss: -2084.464063\n",
            "Train Epoch: 95 [25600/60000 (43%)]\tLoss: -2009.904883\n",
            "Train Epoch: 95 [26880/60000 (45%)]\tLoss: -2005.495312\n",
            "Train Epoch: 95 [28160/60000 (47%)]\tLoss: -2090.278516\n",
            "Train Epoch: 95 [29440/60000 (49%)]\tLoss: -1959.248437\n",
            "Train Epoch: 95 [30720/60000 (51%)]\tLoss: -1968.396875\n",
            "Train Epoch: 95 [32000/60000 (53%)]\tLoss: -2084.945312\n",
            "Train Epoch: 95 [33280/60000 (55%)]\tLoss: -2028.164258\n",
            "Train Epoch: 95 [34560/60000 (58%)]\tLoss: -2002.178906\n",
            "Train Epoch: 95 [35840/60000 (60%)]\tLoss: -1995.088281\n",
            "Train Epoch: 95 [37120/60000 (62%)]\tLoss: -2078.361523\n",
            "Train Epoch: 95 [38400/60000 (64%)]\tLoss: -1987.426953\n",
            "Train Epoch: 95 [39680/60000 (66%)]\tLoss: -2076.305469\n",
            "Train Epoch: 95 [40960/60000 (68%)]\tLoss: -2080.798047\n",
            "Train Epoch: 95 [42240/60000 (70%)]\tLoss: -2066.350391\n",
            "Train Epoch: 95 [43520/60000 (72%)]\tLoss: -2042.344727\n",
            "Train Epoch: 95 [44800/60000 (75%)]\tLoss: -2062.705273\n",
            "Train Epoch: 95 [46080/60000 (77%)]\tLoss: -2076.754102\n",
            "Train Epoch: 95 [47360/60000 (79%)]\tLoss: -2093.721680\n",
            "Train Epoch: 95 [48640/60000 (81%)]\tLoss: -2101.179492\n",
            "Train Epoch: 95 [49920/60000 (83%)]\tLoss: -2006.139258\n",
            "Train Epoch: 95 [51200/60000 (85%)]\tLoss: -2239.935156\n",
            "Train Epoch: 95 [52480/60000 (87%)]\tLoss: -1985.040234\n",
            "Train Epoch: 95 [53760/60000 (90%)]\tLoss: -1990.098828\n",
            "Train Epoch: 95 [55040/60000 (92%)]\tLoss: -1980.825000\n",
            "Train Epoch: 95 [56320/60000 (94%)]\tLoss: -1944.600000\n",
            "Train Epoch: 95 [57600/60000 (96%)]\tLoss: -1993.487891\n",
            "Train Epoch: 95 [58880/60000 (98%)]\tLoss: -2028.250781\n",
            "====> Epoch: 95 Average loss: -20586.6134\n",
            "Train Epoch: 96 [0/60000 (0%)]\tLoss: -20301.673828\n",
            "Train Epoch: 96 [1280/60000 (2%)]\tLoss: -2214.851953\n",
            "Train Epoch: 96 [2560/60000 (4%)]\tLoss: -1939.404883\n",
            "Train Epoch: 96 [3840/60000 (6%)]\tLoss: -2008.660938\n",
            "Train Epoch: 96 [5120/60000 (9%)]\tLoss: -2266.613672\n",
            "Train Epoch: 96 [6400/60000 (11%)]\tLoss: -2040.622266\n",
            "Train Epoch: 96 [7680/60000 (13%)]\tLoss: -2035.058789\n",
            "Train Epoch: 96 [8960/60000 (15%)]\tLoss: -2065.813086\n",
            "Train Epoch: 96 [10240/60000 (17%)]\tLoss: -2314.421289\n",
            "Train Epoch: 96 [11520/60000 (19%)]\tLoss: -2027.675977\n",
            "Train Epoch: 96 [12800/60000 (21%)]\tLoss: -2058.398242\n",
            "Train Epoch: 96 [14080/60000 (23%)]\tLoss: -2396.101562\n",
            "Train Epoch: 96 [15360/60000 (26%)]\tLoss: -1998.789648\n",
            "Train Epoch: 96 [16640/60000 (28%)]\tLoss: -1942.021875\n",
            "Train Epoch: 96 [17920/60000 (30%)]\tLoss: -1995.320312\n",
            "Train Epoch: 96 [19200/60000 (32%)]\tLoss: -2155.740430\n",
            "Train Epoch: 96 [20480/60000 (34%)]\tLoss: -2038.781641\n",
            "Train Epoch: 96 [21760/60000 (36%)]\tLoss: -2283.484961\n",
            "Train Epoch: 96 [23040/60000 (38%)]\tLoss: -2137.666211\n",
            "Train Epoch: 96 [24320/60000 (41%)]\tLoss: -2111.242969\n",
            "Train Epoch: 96 [25600/60000 (43%)]\tLoss: -2025.513281\n",
            "Train Epoch: 96 [26880/60000 (45%)]\tLoss: -2015.651367\n",
            "Train Epoch: 96 [28160/60000 (47%)]\tLoss: -2085.784180\n",
            "Train Epoch: 96 [29440/60000 (49%)]\tLoss: -1975.762109\n",
            "Train Epoch: 96 [30720/60000 (51%)]\tLoss: -1962.615820\n",
            "Train Epoch: 96 [32000/60000 (53%)]\tLoss: -2043.806836\n",
            "Train Epoch: 96 [33280/60000 (55%)]\tLoss: -2038.136719\n",
            "Train Epoch: 96 [34560/60000 (58%)]\tLoss: -2009.233594\n",
            "Train Epoch: 96 [35840/60000 (60%)]\tLoss: -2005.576367\n",
            "Train Epoch: 96 [37120/60000 (62%)]\tLoss: -2115.534570\n",
            "Train Epoch: 96 [38400/60000 (64%)]\tLoss: -2016.489648\n",
            "Train Epoch: 96 [39680/60000 (66%)]\tLoss: -2052.967969\n",
            "Train Epoch: 96 [40960/60000 (68%)]\tLoss: -2059.843945\n",
            "Train Epoch: 96 [42240/60000 (70%)]\tLoss: -2112.822656\n",
            "Train Epoch: 96 [43520/60000 (72%)]\tLoss: -2072.431055\n",
            "Train Epoch: 96 [44800/60000 (75%)]\tLoss: -2078.640039\n",
            "Train Epoch: 96 [46080/60000 (77%)]\tLoss: -2061.806055\n",
            "Train Epoch: 96 [47360/60000 (79%)]\tLoss: -2113.316016\n",
            "Train Epoch: 96 [48640/60000 (81%)]\tLoss: -2126.297461\n",
            "Train Epoch: 96 [49920/60000 (83%)]\tLoss: -2022.482812\n",
            "Train Epoch: 96 [51200/60000 (85%)]\tLoss: -2232.340039\n",
            "Train Epoch: 96 [52480/60000 (87%)]\tLoss: -1974.736328\n",
            "Train Epoch: 96 [53760/60000 (90%)]\tLoss: -2004.360547\n",
            "Train Epoch: 96 [55040/60000 (92%)]\tLoss: -2005.200195\n",
            "Train Epoch: 96 [56320/60000 (94%)]\tLoss: -1942.276172\n",
            "Train Epoch: 96 [57600/60000 (96%)]\tLoss: -1984.957227\n",
            "Train Epoch: 96 [58880/60000 (98%)]\tLoss: -2062.204297\n",
            "====> Epoch: 96 Average loss: -20635.8944\n",
            "Train Epoch: 97 [0/60000 (0%)]\tLoss: -20984.222656\n",
            "Train Epoch: 97 [1280/60000 (2%)]\tLoss: -2223.244141\n",
            "Train Epoch: 97 [2560/60000 (4%)]\tLoss: -2008.240625\n",
            "Train Epoch: 97 [3840/60000 (6%)]\tLoss: -2048.806250\n",
            "Train Epoch: 97 [5120/60000 (9%)]\tLoss: -2225.947656\n",
            "Train Epoch: 97 [6400/60000 (11%)]\tLoss: -2037.031055\n",
            "Train Epoch: 97 [7680/60000 (13%)]\tLoss: -1984.552734\n",
            "Train Epoch: 97 [8960/60000 (15%)]\tLoss: -2037.135352\n",
            "Train Epoch: 97 [10240/60000 (17%)]\tLoss: -2286.343750\n",
            "Train Epoch: 97 [11520/60000 (19%)]\tLoss: -2078.782813\n",
            "Train Epoch: 97 [12800/60000 (21%)]\tLoss: -2056.033984\n",
            "Train Epoch: 97 [14080/60000 (23%)]\tLoss: -2353.385156\n",
            "Train Epoch: 97 [15360/60000 (26%)]\tLoss: -2004.126172\n",
            "Train Epoch: 97 [16640/60000 (28%)]\tLoss: -1953.736523\n",
            "Train Epoch: 97 [17920/60000 (30%)]\tLoss: -2064.572070\n",
            "Train Epoch: 97 [19200/60000 (32%)]\tLoss: -2179.769336\n",
            "Train Epoch: 97 [20480/60000 (34%)]\tLoss: -2001.427734\n",
            "Train Epoch: 97 [21760/60000 (36%)]\tLoss: -2296.732031\n",
            "Train Epoch: 97 [23040/60000 (38%)]\tLoss: -2156.678906\n",
            "Train Epoch: 97 [24320/60000 (41%)]\tLoss: -2078.363672\n",
            "Train Epoch: 97 [25600/60000 (43%)]\tLoss: -2044.590625\n",
            "Train Epoch: 97 [26880/60000 (45%)]\tLoss: -2001.486523\n",
            "Train Epoch: 97 [28160/60000 (47%)]\tLoss: -2091.971875\n",
            "Train Epoch: 97 [29440/60000 (49%)]\tLoss: -1976.629102\n",
            "Train Epoch: 97 [30720/60000 (51%)]\tLoss: -1987.987305\n",
            "Train Epoch: 97 [32000/60000 (53%)]\tLoss: -2060.948047\n",
            "Train Epoch: 97 [33280/60000 (55%)]\tLoss: -2025.327930\n",
            "Train Epoch: 97 [34560/60000 (58%)]\tLoss: -2029.040820\n",
            "Train Epoch: 97 [35840/60000 (60%)]\tLoss: -1999.484766\n",
            "Train Epoch: 97 [37120/60000 (62%)]\tLoss: -2176.122266\n",
            "Train Epoch: 97 [38400/60000 (64%)]\tLoss: -2016.849805\n",
            "Train Epoch: 97 [39680/60000 (66%)]\tLoss: -1994.333594\n",
            "Train Epoch: 97 [40960/60000 (68%)]\tLoss: -2048.048242\n",
            "Train Epoch: 97 [42240/60000 (70%)]\tLoss: -2121.244922\n",
            "Train Epoch: 97 [43520/60000 (72%)]\tLoss: -2147.471484\n",
            "Train Epoch: 97 [44800/60000 (75%)]\tLoss: -2060.724023\n",
            "Train Epoch: 97 [46080/60000 (77%)]\tLoss: -2073.403125\n",
            "Train Epoch: 97 [47360/60000 (79%)]\tLoss: -2057.456836\n",
            "Train Epoch: 97 [48640/60000 (81%)]\tLoss: -2097.683984\n",
            "Train Epoch: 97 [49920/60000 (83%)]\tLoss: -2005.120508\n",
            "Train Epoch: 97 [51200/60000 (85%)]\tLoss: -2245.977930\n",
            "Train Epoch: 97 [52480/60000 (87%)]\tLoss: -1999.003516\n",
            "Train Epoch: 97 [53760/60000 (90%)]\tLoss: -2006.842773\n",
            "Train Epoch: 97 [55040/60000 (92%)]\tLoss: -2003.193555\n",
            "Train Epoch: 97 [56320/60000 (94%)]\tLoss: -1951.498437\n",
            "Train Epoch: 97 [57600/60000 (96%)]\tLoss: -1988.958008\n",
            "Train Epoch: 97 [58880/60000 (98%)]\tLoss: -1982.131250\n",
            "====> Epoch: 97 Average loss: -20670.4547\n",
            "Train Epoch: 98 [0/60000 (0%)]\tLoss: -20867.007812\n",
            "Train Epoch: 98 [1280/60000 (2%)]\tLoss: -2261.189258\n",
            "Train Epoch: 98 [2560/60000 (4%)]\tLoss: -2036.408008\n",
            "Train Epoch: 98 [3840/60000 (6%)]\tLoss: -2057.591016\n",
            "Train Epoch: 98 [5120/60000 (9%)]\tLoss: -2180.494336\n",
            "Train Epoch: 98 [6400/60000 (11%)]\tLoss: -2026.447852\n",
            "Train Epoch: 98 [7680/60000 (13%)]\tLoss: -2004.651172\n",
            "Train Epoch: 98 [8960/60000 (15%)]\tLoss: -2045.140430\n",
            "Train Epoch: 98 [10240/60000 (17%)]\tLoss: -2249.009375\n",
            "Train Epoch: 98 [11520/60000 (19%)]\tLoss: -2067.132227\n",
            "Train Epoch: 98 [12800/60000 (21%)]\tLoss: -2052.811914\n",
            "Train Epoch: 98 [14080/60000 (23%)]\tLoss: -2265.886914\n",
            "Train Epoch: 98 [15360/60000 (26%)]\tLoss: -1990.158984\n",
            "Train Epoch: 98 [16640/60000 (28%)]\tLoss: -1961.895703\n",
            "Train Epoch: 98 [17920/60000 (30%)]\tLoss: -2069.861328\n",
            "Train Epoch: 98 [19200/60000 (32%)]\tLoss: -2144.201367\n",
            "Train Epoch: 98 [20480/60000 (34%)]\tLoss: -2006.875586\n",
            "Train Epoch: 98 [21760/60000 (36%)]\tLoss: -2307.228125\n",
            "Train Epoch: 98 [23040/60000 (38%)]\tLoss: -2205.836914\n",
            "Train Epoch: 98 [24320/60000 (41%)]\tLoss: -2108.465234\n",
            "Train Epoch: 98 [25600/60000 (43%)]\tLoss: -2018.128906\n",
            "Train Epoch: 98 [26880/60000 (45%)]\tLoss: -2007.753516\n",
            "Train Epoch: 98 [28160/60000 (47%)]\tLoss: -2182.526953\n",
            "Train Epoch: 98 [29440/60000 (49%)]\tLoss: -1970.746484\n",
            "Train Epoch: 98 [30720/60000 (51%)]\tLoss: -1983.292383\n",
            "Train Epoch: 98 [32000/60000 (53%)]\tLoss: -2095.886133\n",
            "Train Epoch: 98 [33280/60000 (55%)]\tLoss: -2016.163086\n",
            "Train Epoch: 98 [34560/60000 (58%)]\tLoss: -1999.792383\n",
            "Train Epoch: 98 [35840/60000 (60%)]\tLoss: -1987.426367\n",
            "Train Epoch: 98 [37120/60000 (62%)]\tLoss: -2205.676758\n",
            "Train Epoch: 98 [38400/60000 (64%)]\tLoss: -2031.635547\n",
            "Train Epoch: 98 [39680/60000 (66%)]\tLoss: -1992.558398\n",
            "Train Epoch: 98 [40960/60000 (68%)]\tLoss: -2051.133594\n",
            "Train Epoch: 98 [42240/60000 (70%)]\tLoss: -2162.837500\n",
            "Train Epoch: 98 [43520/60000 (72%)]\tLoss: -2165.869141\n",
            "Train Epoch: 98 [44800/60000 (75%)]\tLoss: -2083.739648\n",
            "Train Epoch: 98 [46080/60000 (77%)]\tLoss: -2099.454297\n",
            "Train Epoch: 98 [47360/60000 (79%)]\tLoss: -2113.416016\n",
            "Train Epoch: 98 [48640/60000 (81%)]\tLoss: -2097.688281\n",
            "Train Epoch: 98 [49920/60000 (83%)]\tLoss: -2010.938867\n",
            "Train Epoch: 98 [51200/60000 (85%)]\tLoss: -2240.634180\n",
            "Train Epoch: 98 [52480/60000 (87%)]\tLoss: -2050.947070\n",
            "Train Epoch: 98 [53760/60000 (90%)]\tLoss: -2031.133594\n",
            "Train Epoch: 98 [55040/60000 (92%)]\tLoss: -2005.944922\n",
            "Train Epoch: 98 [56320/60000 (94%)]\tLoss: -1968.180273\n",
            "Train Epoch: 98 [57600/60000 (96%)]\tLoss: -1995.287500\n",
            "Train Epoch: 98 [58880/60000 (98%)]\tLoss: -1966.562695\n",
            "====> Epoch: 98 Average loss: -20674.5743\n",
            "Train Epoch: 99 [0/60000 (0%)]\tLoss: -20623.228516\n",
            "Train Epoch: 99 [1280/60000 (2%)]\tLoss: -2246.692383\n",
            "Train Epoch: 99 [2560/60000 (4%)]\tLoss: -2029.087305\n",
            "Train Epoch: 99 [3840/60000 (6%)]\tLoss: -2035.929492\n",
            "Train Epoch: 99 [5120/60000 (9%)]\tLoss: -2252.839063\n",
            "Train Epoch: 99 [6400/60000 (11%)]\tLoss: -2026.699219\n",
            "Train Epoch: 99 [7680/60000 (13%)]\tLoss: -2028.386133\n",
            "Train Epoch: 99 [8960/60000 (15%)]\tLoss: -2037.682227\n",
            "Train Epoch: 99 [10240/60000 (17%)]\tLoss: -2237.601953\n",
            "Train Epoch: 99 [11520/60000 (19%)]\tLoss: -2053.160352\n",
            "Train Epoch: 99 [12800/60000 (21%)]\tLoss: -2074.896094\n",
            "Train Epoch: 99 [14080/60000 (23%)]\tLoss: -2261.281641\n",
            "Train Epoch: 99 [15360/60000 (26%)]\tLoss: -2015.142969\n",
            "Train Epoch: 99 [16640/60000 (28%)]\tLoss: -1983.525977\n",
            "Train Epoch: 99 [17920/60000 (30%)]\tLoss: -2077.946875\n",
            "Train Epoch: 99 [19200/60000 (32%)]\tLoss: -2134.842773\n",
            "Train Epoch: 99 [20480/60000 (34%)]\tLoss: -2017.829883\n",
            "Train Epoch: 99 [21760/60000 (36%)]\tLoss: -2288.085156\n",
            "Train Epoch: 99 [23040/60000 (38%)]\tLoss: -2186.359766\n",
            "Train Epoch: 99 [24320/60000 (41%)]\tLoss: -2085.123047\n",
            "Train Epoch: 99 [25600/60000 (43%)]\tLoss: -2007.400000\n",
            "Train Epoch: 99 [26880/60000 (45%)]\tLoss: -2003.886523\n",
            "Train Epoch: 99 [28160/60000 (47%)]\tLoss: -2133.431641\n",
            "Train Epoch: 99 [29440/60000 (49%)]\tLoss: -1989.241797\n",
            "Train Epoch: 99 [30720/60000 (51%)]\tLoss: -1976.735547\n",
            "Train Epoch: 99 [32000/60000 (53%)]\tLoss: -2081.719531\n",
            "Train Epoch: 99 [33280/60000 (55%)]\tLoss: -1981.834766\n",
            "Train Epoch: 99 [34560/60000 (58%)]\tLoss: -1962.692188\n",
            "Train Epoch: 99 [35840/60000 (60%)]\tLoss: -1984.596094\n",
            "Train Epoch: 99 [37120/60000 (62%)]\tLoss: -2163.521680\n",
            "Train Epoch: 99 [38400/60000 (64%)]\tLoss: -2014.700000\n",
            "Train Epoch: 99 [39680/60000 (66%)]\tLoss: -2001.610352\n",
            "Train Epoch: 99 [40960/60000 (68%)]\tLoss: -2042.674609\n",
            "Train Epoch: 99 [42240/60000 (70%)]\tLoss: -2136.060937\n",
            "Train Epoch: 99 [43520/60000 (72%)]\tLoss: -2177.066797\n",
            "Train Epoch: 99 [44800/60000 (75%)]\tLoss: -2049.803906\n",
            "Train Epoch: 99 [46080/60000 (77%)]\tLoss: -2079.651953\n",
            "Train Epoch: 99 [47360/60000 (79%)]\tLoss: -2121.445703\n",
            "Train Epoch: 99 [48640/60000 (81%)]\tLoss: -2078.281250\n",
            "Train Epoch: 99 [49920/60000 (83%)]\tLoss: -2012.009375\n",
            "Train Epoch: 99 [51200/60000 (85%)]\tLoss: -2223.742383\n",
            "Train Epoch: 99 [52480/60000 (87%)]\tLoss: -1997.720117\n",
            "Train Epoch: 99 [53760/60000 (90%)]\tLoss: -2050.354297\n",
            "Train Epoch: 99 [55040/60000 (92%)]\tLoss: -2019.267578\n",
            "Train Epoch: 99 [56320/60000 (94%)]\tLoss: -1959.007617\n",
            "Train Epoch: 99 [57600/60000 (96%)]\tLoss: -1996.394336\n",
            "Train Epoch: 99 [58880/60000 (98%)]\tLoss: -1962.120508\n",
            "====> Epoch: 99 Average loss: -20668.4369\n",
            "Train Epoch: 100 [0/60000 (0%)]\tLoss: -20350.626953\n",
            "Train Epoch: 100 [1280/60000 (2%)]\tLoss: -2214.000000\n",
            "Train Epoch: 100 [2560/60000 (4%)]\tLoss: -2031.940039\n",
            "Train Epoch: 100 [3840/60000 (6%)]\tLoss: -2050.336133\n",
            "Train Epoch: 100 [5120/60000 (9%)]\tLoss: -2203.530469\n",
            "Train Epoch: 100 [6400/60000 (11%)]\tLoss: -2037.038477\n",
            "Train Epoch: 100 [7680/60000 (13%)]\tLoss: -2035.500781\n",
            "Train Epoch: 100 [8960/60000 (15%)]\tLoss: -2038.126953\n",
            "Train Epoch: 100 [10240/60000 (17%)]\tLoss: -2287.633594\n",
            "Train Epoch: 100 [11520/60000 (19%)]\tLoss: -2042.147656\n",
            "Train Epoch: 100 [12800/60000 (21%)]\tLoss: -2062.563281\n",
            "Train Epoch: 100 [14080/60000 (23%)]\tLoss: -2277.468164\n",
            "Train Epoch: 100 [15360/60000 (26%)]\tLoss: -2035.516211\n",
            "Train Epoch: 100 [16640/60000 (28%)]\tLoss: -1958.048047\n",
            "Train Epoch: 100 [17920/60000 (30%)]\tLoss: -2062.431250\n",
            "Train Epoch: 100 [19200/60000 (32%)]\tLoss: -2163.401172\n",
            "Train Epoch: 100 [20480/60000 (34%)]\tLoss: -2004.677734\n",
            "Train Epoch: 100 [21760/60000 (36%)]\tLoss: -2286.041797\n",
            "Train Epoch: 100 [23040/60000 (38%)]\tLoss: -2212.975000\n",
            "Train Epoch: 100 [24320/60000 (41%)]\tLoss: -2073.207227\n",
            "Train Epoch: 100 [25600/60000 (43%)]\tLoss: -1992.887500\n",
            "Train Epoch: 100 [26880/60000 (45%)]\tLoss: -2006.196680\n",
            "Train Epoch: 100 [28160/60000 (47%)]\tLoss: -2121.083398\n",
            "Train Epoch: 100 [29440/60000 (49%)]\tLoss: -1998.592969\n",
            "Train Epoch: 100 [30720/60000 (51%)]\tLoss: -1992.769727\n",
            "Train Epoch: 100 [32000/60000 (53%)]\tLoss: -2079.869141\n",
            "Train Epoch: 100 [33280/60000 (55%)]\tLoss: -2000.174023\n",
            "Train Epoch: 100 [34560/60000 (58%)]\tLoss: -1973.532031\n",
            "Train Epoch: 100 [35840/60000 (60%)]\tLoss: -1983.128711\n",
            "Train Epoch: 100 [37120/60000 (62%)]\tLoss: -2172.185352\n",
            "Train Epoch: 100 [38400/60000 (64%)]\tLoss: -2000.652734\n",
            "Train Epoch: 100 [39680/60000 (66%)]\tLoss: -1974.009180\n",
            "Train Epoch: 100 [40960/60000 (68%)]\tLoss: -2060.187109\n",
            "Train Epoch: 100 [42240/60000 (70%)]\tLoss: -2156.984180\n",
            "Train Epoch: 100 [43520/60000 (72%)]\tLoss: -2158.318359\n",
            "Train Epoch: 100 [44800/60000 (75%)]\tLoss: -2051.888867\n",
            "Train Epoch: 100 [46080/60000 (77%)]\tLoss: -2094.815820\n",
            "Train Epoch: 100 [47360/60000 (79%)]\tLoss: -2135.891406\n",
            "Train Epoch: 100 [48640/60000 (81%)]\tLoss: -2071.249609\n",
            "Train Epoch: 100 [49920/60000 (83%)]\tLoss: -2009.751367\n",
            "Train Epoch: 100 [51200/60000 (85%)]\tLoss: -2249.987305\n",
            "Train Epoch: 100 [52480/60000 (87%)]\tLoss: -2014.695117\n",
            "Train Epoch: 100 [53760/60000 (90%)]\tLoss: -2006.073828\n",
            "Train Epoch: 100 [55040/60000 (92%)]\tLoss: -2027.978320\n",
            "Train Epoch: 100 [56320/60000 (94%)]\tLoss: -1952.846094\n",
            "Train Epoch: 100 [57600/60000 (96%)]\tLoss: -1991.728516\n",
            "Train Epoch: 100 [58880/60000 (98%)]\tLoss: -2001.275000\n",
            "====> Epoch: 100 Average loss: -20660.6135\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l1WpIURTaqOi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# !rm -rf \"./cvae/models/Epoch*\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gTIIsTNHYn38",
        "colab_type": "text"
      },
      "source": [
        "See the model output that compare the original image with generated image"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ODhnG1JvBBVQ",
        "colab_type": "code",
        "outputId": "7b2f4ac5-e0c1-4d63-c2c5-c48477d0abf7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        }
      },
      "source": [
        "from torch.utils.data import RandomSampler\n",
        "from torchvision.utils import save_image\n",
        "import matplotlib.image as mpimg\n",
        "\n",
        "with torch.no_grad():\n",
        "  test_loader = DataLoader(mnist_data,batch_size=16, shuffle=True)\n",
        "  itr = iter(test_loader)\n",
        "  # next(itr)\n",
        "  batch, label = next(itr)\n",
        "  one_hot = get_one_hot(len(label), label)\n",
        "  batch = batch.to(device)\n",
        "  model.eval()\n",
        "  recon_batch, _, _  = model(batch, one_hot)\n",
        "\n",
        "  show(batch.cpu(), renorm = True, nrow=4)\n",
        "  show(recon_batch.cpu(), renorm = False, nrow=4)\n"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMceM4v8BihChG1G7CdUAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOcAAADnCAYAAADl9EEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\naFP8Hx+Hr2SxpAB0AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3_ALyF0qQWWd",
        "colab_type": "code",
        "outputId": "38aff31d-db0b-4d04-81be-c3d3a2dd4b8d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "source": [
        "with torch.no_grad():\n",
        "    epoch = 1\n",
        "    rand1 = torch.randn(9, latent_dim).to(device)\n",
        "    rand2 = torch.randn(9, latent_dim).to(device)\n",
        "\n",
        "    z = model.reparameterize(rand1, rand2)\n",
        "    labels = torch.from_numpy(np.arange(1,10))\n",
        "    print(labels)\n",
        "    one_hot = get_one_hot(len(labels), labels)\n",
        "    sample = model.decode(z, one_hot).cpu()\n",
        "    print(sample.shape)\n",
        "    sample = sample.view(9, 1, 28, 28)\n",
        "    \n",
        "    show(sample, renorm = False, nrow=9)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([1, 2, 3, 4, 5, 6, 7, 8, 9])\n",
            "torch.Size([9, 784])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAA1CAYAAADiWDg0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2zO60_SdZimP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p5Bmxbt2Zkwk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!cp -rf \"./cvae\" \"drive/My Drive/Colab Notebooks/\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xS2QkAKTZmYb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!cp -rf \"drive/My Drive/Colab Notebooks/cvae\" \"./cvae\" "
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}