From 7644bdacf49ccb36bca219b5f7e7860cbf9d2428 Mon Sep 17 00:00:00 2001
From: shlear <116897538+shlear@users.noreply.github.com>
Date: Mon, 30 Jan 2023 09:57:08 +0300
Subject: [PATCH] =?UTF-8?q?=D0=A1=D0=BE=D0=B7=D0=B4=D0=B0=D0=BD=D0=BE=20?=
 =?UTF-8?q?=D1=81=20=D0=BF=D0=BE=D0=BC=D0=BE=D1=89=D1=8C=D1=8E=20Colaborat?=
 =?UTF-8?q?ory?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 13-GANs/GAN_HW.ipynb | 3815 ++++++++++++++++++++++++++++++++++++++----
 1 file changed, 3499 insertions(+), 316 deletions(-)

diff --git a/13-GANs/GAN_HW.ipynb b/13-GANs/GAN_HW.ipynb
index c4bef55..a6d0c20 100644
--- a/13-GANs/GAN_HW.ipynb
+++ b/13-GANs/GAN_HW.ipynb
@@ -1,318 +1,3501 @@
 {
- "cells": [
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "<a href=\"https://colab.research.google.com/github/HSE-LAMBDA/MLDM-2022/blob/main/13-GANs/GAN_HW.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
-   ]
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "p0Kyda8zUdro"
+      },
+      "source": [
+        "<a href=\"https://colab.research.google.com/github/HSE-LAMBDA/MLDM-2022/blob/main/13-GANs/GAN_HW.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "Eaip0Gm0Udrq"
+      },
+      "source": [
+        "### WGAN\n",
+        "\n",
+        "* Modify snippets below and implement [Wasserstein GAN](https://arxiv.org/abs/1701.07875) with weight clipping. (2 points)\n",
+        "\n",
+        "* Replace  weight clipping with [gradient penalty](https://arxiv.org/pdf/1704.00028v3.pdf). (2 points)\n",
+        "\n",
+        "* Add labels into WGAN, performing [conditional generation](https://arxiv.org/pdf/1411.1784.pdf). (2 points) \n",
+        "\n",
+        "Write a report about experiments and results, add plots and visualizations."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 1,
+      "metadata": {
+        "id": "aBack-CyUdrq"
+      },
+      "outputs": [],
+      "source": [
+        "import torch\n",
+        "import torch.nn as nn\n",
+        "import torch.nn.functional as F\n",
+        "import torch.optim as optim\n",
+        "from torch.utils.data import DataLoader, Dataset\n",
+        "\n",
+        "import torchvision\n",
+        "import matplotlib.pyplot as plt\n",
+        "import numpy as np\n",
+        "\n",
+        "from torch.autograd import Variable"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "SD7I7vZ8Udrr"
+      },
+      "source": [
+        "### Creating config object (argparse workaround)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 2,
+      "metadata": {
+        "id": "e0Yq4aMxUdrr"
+      },
+      "outputs": [],
+      "source": [
+        "class Config:\n",
+        "    pass\n",
+        "\n",
+        "config = Config()\n",
+        "config.mnist_path = None\n",
+        "config.batch_size = 16\n",
+        "config.num_workers = 3\n",
+        "config.num_epochs = 10\n",
+        "config.noise_size = 50\n",
+        "config.print_freq = 100\n"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "2CGmqp35Udrr"
+      },
+      "source": [
+        "### Create dataloder"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 3,
+      "metadata": {
+        "id": "jtK105iuUdrr",
+        "outputId": "9e43719a-08a1-4bfc-e426-f89fd4019cf4",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 451,
+          "referenced_widgets": [
+            "3089611b5cda418faf550134695bd1c2",
+            "02e561b7133b4c5e920a3a2c0d300d2e",
+            "d25fd1b9a05f404eb37505b593c83bee",
+            "70d8400cd7474cf3984d1f31eb3f6fd7",
+            "ab3ec40e6808496dbb5e46d77773eff7",
+            "993857a0d4574ff8bef19a791c0b9e3c",
+            "00258c5fbf8348a79f9d859a34a5df43",
+            "37d0caea6f5146c98eb549c98b113d9a",
+            "26a20977835d4f8ea9dbc0f06427bdbb",
+            "8c987b2fd7c249e7b74623c91a3f15ce",
+            "fb3036cf1cd0437b86ac1632219f7e04",
+            "65be345916ef419893271ff1f0a92b51",
+            "37c36bd250e04124a09270e46141f41f",
+            "2937ae2f576b40ada1f51eac29d28422",
+            "f68051dc0ef54f0f978d67f303a22e9d",
+            "dd0cbf60b58149b89f0cd0e59c130b0e",
+            "7a2f922f6aae441380f1586048914365",
+            "933951cefdc84b7796526bdc6869fdf3",
+            "3ff4afbc672c40d0b096f119a07d0024",
+            "b02746fe4d774ca39040f7685f7827cf",
+            "701dc00c6a24482192ecf3f082babc2c",
+            "c6547b6dabf148508e341c2bdf369d83",
+            "92cefc1774de486091dc3b60021783c1",
+            "8fa40adeb3cb428298430949d13ad090",
+            "c3ab1c12ac574a15920daf69c738c302",
+            "5a8a21e352be4118a88e75141ac843cc",
+            "37b2250444864bf880421d4d2fdc57ae",
+            "2d9f91fe44054b9ca60f153878eddcba",
+            "f271a7fe389546edbeddc0cee11444f4",
+            "ba1d248606e54691a73a062a9f5646cb",
+            "340311831d5b492e99c0a5e015975443",
+            "2f0bf104af174a21aecc2a320cece5c5",
+            "3bfc8f6bf8a6430693e0703b22a57272",
+            "e251c830789e469888cc2173200a66ff",
+            "a6c11c59a074439cb9417b0b3047ff37",
+            "4e0ce38999fd431c971f095a11493445",
+            "5daff87ee00f4024a4d0c36ec365656d",
+            "89306a31ada645e887f280483191d061",
+            "83b00e40330e4a9bab89fce2d40aab7f",
+            "77b51cbd87e04728a058d46ee733c819",
+            "c5dabb6aa34d4808bfe8a876a15130b4",
+            "643947d50110467fab73bb2609e40467",
+            "fffcb3453216492cbb954f130aeaba29",
+            "7d11aa9d9be3406581f64757a5550af7"
+          ]
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz to fashion_mnist/FashionMNIST/raw/train-images-idx3-ubyte.gz\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "  0%|          | 0/26421880 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "3089611b5cda418faf550134695bd1c2"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Extracting fashion_mnist/FashionMNIST/raw/train-images-idx3-ubyte.gz to fashion_mnist/FashionMNIST/raw\n",
+            "\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz to fashion_mnist/FashionMNIST/raw/train-labels-idx1-ubyte.gz\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "  0%|          | 0/29515 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "65be345916ef419893271ff1f0a92b51"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Extracting fashion_mnist/FashionMNIST/raw/train-labels-idx1-ubyte.gz to fashion_mnist/FashionMNIST/raw\n",
+            "\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz to fashion_mnist/FashionMNIST/raw/t10k-images-idx3-ubyte.gz\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "  0%|          | 0/4422102 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "92cefc1774de486091dc3b60021783c1"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Extracting fashion_mnist/FashionMNIST/raw/t10k-images-idx3-ubyte.gz to fashion_mnist/FashionMNIST/raw\n",
+            "\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz\n",
+            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz to fashion_mnist/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "  0%|          | 0/5148 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "e251c830789e469888cc2173200a66ff"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Extracting fashion_mnist/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz to fashion_mnist/FashionMNIST/raw\n",
+            "\n"
+          ]
+        }
+      ],
+      "source": [
+        "train = torchvision.datasets.FashionMNIST(\"fashion_mnist\", train=True, transform=torchvision.transforms.ToTensor(), download=True)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 4,
+      "metadata": {
+        "id": "ghEDU08xUdrs"
+      },
+      "outputs": [],
+      "source": [
+        "dataloader = DataLoader(train, batch_size=16, shuffle=True)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 5,
+      "metadata": {
+        "id": "AlyUytopUdrs",
+        "outputId": "b999ea90-080e-4d4b-c872-61a4775fcbe9",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "3750"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 5
+        }
+      ],
+      "source": [
+        "len(dataloader)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 6,
+      "metadata": {
+        "id": "ai4GQlhUUdrs"
+      },
+      "outputs": [],
+      "source": [
+        "for image, cat in dataloader:\n",
+        "    break"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 7,
+      "metadata": {
+        "scrolled": true,
+        "id": "AHjavRq0Udrs",
+        "outputId": "10060bac-561d-4557-cbe8-430a56226331",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "torch.Size([16, 1, 28, 28])"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 7
+        }
+      ],
+      "source": [
+        "image.size()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "ltR36cKpUdrs"
+      },
+      "source": [
+        "### Create generator and discriminator"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 8,
+      "metadata": {
+        "id": "FTxWWtRZUdrs"
+      },
+      "outputs": [],
+      "source": [
+        "class Generator(nn.Module):\n",
+        "    def __init__(self):\n",
+        "        super(Generator, self).__init__()\n",
+        "        self.model = nn.Sequential( \n",
+        "            nn.Linear(config.noise_size, 200),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(200, 28*28),\n",
+        "            nn.Sigmoid())\n",
+        "        \n",
+        "    def forward(self, x):\n",
+        "        return self.model(x)\n",
+        "    \n",
+        "class Discriminator(nn.Module):\n",
+        "    def __init__(self):\n",
+        "        super(Discriminator, self).__init__()\n",
+        "        self.model = nn.Sequential(\n",
+        "            nn.Linear(28*28, 200),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(200, 50),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(50, 1), \n",
+        "            nn.Sigmoid())\n",
+        "    def forward(self, x):\n",
+        "        return self.model(x)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 9,
+      "metadata": {
+        "id": "sdSFPJRGUdrt"
+      },
+      "outputs": [],
+      "source": [
+        "generator = Generator()\n",
+        "discriminator = Discriminator()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "p6mS17QjUdrt"
+      },
+      "source": [
+        "### Create optimizers and loss"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 10,
+      "metadata": {
+        "id": "1i4GrgO1Udrt"
+      },
+      "outputs": [],
+      "source": [
+        "# modification2: use RMSprop (or SGD) instead of Adam\n",
+        "# optimizer\n",
+        "# optim_G = optim.Adam(params=generator.parameters(), lr=0.0001)\n",
+        "# optim_D = optim.Adam(params=discriminator.parameters(), lr=0.0001)\n",
+        "optim_G = optim.RMSprop(params=generator.parameters(),lr=0.00005)\n",
+        "optim_D = optim.RMSprop(params=discriminator.parameters(),lr=0.00005)\n",
+        "\n",
+        "# modification4: no log in loss\n",
+        "# criterion = nn.BCELoss()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "ohwrLM7tUdrt"
+      },
+      "source": [
+        "### Create necessary variables"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 11,
+      "metadata": {
+        "id": "kYfOh6i4Udrt",
+        "outputId": "5a9c7c42-4b40-454f-f16f-c2661d25e19f",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "Discriminator(\n",
+              "  (model): Sequential(\n",
+              "    (0): Linear(in_features=784, out_features=200, bias=True)\n",
+              "    (1): ReLU(inplace=True)\n",
+              "    (2): Linear(in_features=200, out_features=50, bias=True)\n",
+              "    (3): ReLU(inplace=True)\n",
+              "    (4): Linear(in_features=50, out_features=1, bias=True)\n",
+              "    (5): Sigmoid()\n",
+              "  )\n",
+              ")"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 11
+        }
+      ],
+      "source": [
+        "from torch.utils.tensorboard import SummaryWriter\n",
+        "input = Variable(torch.FloatTensor(config.batch_size, 28*28))\n",
+        "noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size))\n",
+        "fixed_noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size).normal_(0, 1))\n",
+        "label = Variable(torch.FloatTensor(config.batch_size))\n",
+        "real_label = 1\n",
+        "fake_label = 0\n",
+        "\n",
+        "# for tensorboard plotting\n",
+        "writer_real = SummaryWriter(f\"logs/real\")\n",
+        "writer_fake = SummaryWriter(f\"logs/fake\")\n",
+        "step = 0\n",
+        "\n",
+        "generator.train()\n",
+        "discriminator.train()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "WiHkx6rOUdrt"
+      },
+      "source": [
+        "### GAN"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 12,
+      "metadata": {
+        "scrolled": true,
+        "id": "zrHq0fJiUdrt",
+        "outputId": "2f6894e2-cc4e-4fde-e20e-d9744964d099",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Epoch [1/10] Batch 100/3750                   Loss D: -0.0229, loss G: -0.4882\n",
+            "Epoch [1/10] Batch 200/3750                   Loss D: -0.0356, loss G: -0.4740\n",
+            "Epoch [1/10] Batch 300/3750                   Loss D: -0.0353, loss G: -0.4621\n",
+            "Epoch [1/10] Batch 400/3750                   Loss D: -0.0285, loss G: -0.4560\n",
+            "Epoch [1/10] Batch 500/3750                   Loss D: -0.0326, loss G: -0.4509\n",
+            "Epoch [1/10] Batch 600/3750                   Loss D: -0.0302, loss G: -0.4530\n",
+            "Epoch [1/10] Batch 700/3750                   Loss D: -0.0251, loss G: -0.4669\n",
+            "Epoch [1/10] Batch 800/3750                   Loss D: -0.0194, loss G: -0.4841\n",
+            "Epoch [1/10] Batch 900/3750                   Loss D: -0.0180, loss G: -0.4940\n",
+            "Epoch [1/10] Batch 1000/3750                   Loss D: -0.0293, loss G: -0.4924\n",
+            "Epoch [1/10] Batch 1100/3750                   Loss D: -0.0286, loss G: -0.4947\n",
+            "Epoch [1/10] Batch 1200/3750                   Loss D: -0.0306, loss G: -0.4999\n",
+            "Epoch [1/10] Batch 1300/3750                   Loss D: -0.0346, loss G: -0.5027\n",
+            "Epoch [1/10] Batch 1400/3750                   Loss D: -0.0189, loss G: -0.5093\n",
+            "Epoch [1/10] Batch 1500/3750                   Loss D: -0.0214, loss G: -0.4972\n",
+            "Epoch [1/10] Batch 1600/3750                   Loss D: -0.0171, loss G: -0.5068\n",
+            "Epoch [1/10] Batch 1700/3750                   Loss D: -0.0230, loss G: -0.5046\n",
+            "Epoch [1/10] Batch 1800/3750                   Loss D: -0.0096, loss G: -0.5098\n",
+            "Epoch [1/10] Batch 1900/3750                   Loss D: -0.0156, loss G: -0.5040\n",
+            "Epoch [1/10] Batch 2000/3750                   Loss D: -0.0223, loss G: -0.4945\n",
+            "Epoch [1/10] Batch 2100/3750                   Loss D: -0.0109, loss G: -0.5107\n",
+            "Epoch [1/10] Batch 2200/3750                   Loss D: -0.0045, loss G: -0.5080\n",
+            "Epoch [1/10] Batch 2300/3750                   Loss D: -0.0066, loss G: -0.5000\n",
+            "Epoch [1/10] Batch 2400/3750                   Loss D: -0.0109, loss G: -0.5068\n",
+            "Epoch [1/10] Batch 2500/3750                   Loss D: -0.0110, loss G: -0.4950\n",
+            "Epoch [1/10] Batch 2600/3750                   Loss D: -0.0054, loss G: -0.5053\n",
+            "Epoch [1/10] Batch 2700/3750                   Loss D: -0.0052, loss G: -0.5074\n",
+            "Epoch [1/10] Batch 2800/3750                   Loss D: -0.0072, loss G: -0.4935\n",
+            "Epoch [1/10] Batch 2900/3750                   Loss D: -0.0040, loss G: -0.4998\n",
+            "Epoch [1/10] Batch 3000/3750                   Loss D: -0.0053, loss G: -0.4981\n",
+            "Epoch [1/10] Batch 3100/3750                   Loss D: -0.0104, loss G: -0.4924\n",
+            "Epoch [1/10] Batch 3200/3750                   Loss D: -0.0045, loss G: -0.4999\n",
+            "Epoch [1/10] Batch 3300/3750                   Loss D: -0.0080, loss G: -0.5013\n",
+            "Epoch [1/10] Batch 3400/3750                   Loss D: -0.0088, loss G: -0.4971\n",
+            "Epoch [1/10] Batch 3500/3750                   Loss D: -0.0090, loss G: -0.4973\n",
+            "Epoch [1/10] Batch 3600/3750                   Loss D: -0.0073, loss G: -0.4991\n",
+            "Epoch [1/10] Batch 3700/3750                   Loss D: -0.0097, loss G: -0.4935\n",
+            "Epoch [2/10] Batch 100/3750                   Loss D: -0.0104, loss G: -0.5051\n",
+            "Epoch [2/10] Batch 200/3750                   Loss D: -0.0113, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 300/3750                   Loss D: -0.0048, loss G: -0.5027\n",
+            "Epoch [2/10] Batch 400/3750                   Loss D: -0.0075, loss G: -0.4984\n",
+            "Epoch [2/10] Batch 500/3750                   Loss D: -0.0089, loss G: -0.4876\n",
+            "Epoch [2/10] Batch 600/3750                   Loss D: -0.0092, loss G: -0.5133\n",
+            "Epoch [2/10] Batch 700/3750                   Loss D: -0.0111, loss G: -0.5033\n",
+            "Epoch [2/10] Batch 800/3750                   Loss D: -0.0101, loss G: -0.4886\n",
+            "Epoch [2/10] Batch 900/3750                   Loss D: -0.0080, loss G: -0.5036\n",
+            "Epoch [2/10] Batch 1000/3750                   Loss D: -0.0106, loss G: -0.5092\n",
+            "Epoch [2/10] Batch 1100/3750                   Loss D: -0.0092, loss G: -0.5003\n",
+            "Epoch [2/10] Batch 1200/3750                   Loss D: -0.0037, loss G: -0.5029\n",
+            "Epoch [2/10] Batch 1300/3750                   Loss D: -0.0092, loss G: -0.5003\n",
+            "Epoch [2/10] Batch 1400/3750                   Loss D: -0.0096, loss G: -0.4990\n",
+            "Epoch [2/10] Batch 1500/3750                   Loss D: -0.0071, loss G: -0.4943\n",
+            "Epoch [2/10] Batch 1600/3750                   Loss D: -0.0065, loss G: -0.5044\n",
+            "Epoch [2/10] Batch 1700/3750                   Loss D: -0.0022, loss G: -0.4976\n",
+            "Epoch [2/10] Batch 1800/3750                   Loss D: -0.0055, loss G: -0.4997\n",
+            "Epoch [2/10] Batch 1900/3750                   Loss D: -0.0024, loss G: -0.5051\n",
+            "Epoch [2/10] Batch 2000/3750                   Loss D: -0.0041, loss G: -0.4902\n",
+            "Epoch [2/10] Batch 2100/3750                   Loss D: 0.0026, loss G: -0.5093\n",
+            "Epoch [2/10] Batch 2200/3750                   Loss D: -0.0048, loss G: -0.4983\n",
+            "Epoch [2/10] Batch 2300/3750                   Loss D: -0.0085, loss G: -0.5017\n",
+            "Epoch [2/10] Batch 2400/3750                   Loss D: -0.0039, loss G: -0.5015\n",
+            "Epoch [2/10] Batch 2500/3750                   Loss D: -0.0025, loss G: -0.5044\n",
+            "Epoch [2/10] Batch 2600/3750                   Loss D: 0.0008, loss G: -0.4867\n",
+            "Epoch [2/10] Batch 2700/3750                   Loss D: -0.0019, loss G: -0.4906\n",
+            "Epoch [2/10] Batch 2800/3750                   Loss D: -0.0019, loss G: -0.5185\n",
+            "Epoch [2/10] Batch 2900/3750                   Loss D: -0.0088, loss G: -0.4897\n",
+            "Epoch [2/10] Batch 3000/3750                   Loss D: -0.0064, loss G: -0.4981\n",
+            "Epoch [2/10] Batch 3100/3750                   Loss D: -0.0044, loss G: -0.5055\n",
+            "Epoch [2/10] Batch 3200/3750                   Loss D: -0.0015, loss G: -0.4973\n",
+            "Epoch [2/10] Batch 3300/3750                   Loss D: -0.0077, loss G: -0.5113\n",
+            "Epoch [2/10] Batch 3400/3750                   Loss D: -0.0034, loss G: -0.4894\n",
+            "Epoch [2/10] Batch 3500/3750                   Loss D: -0.0031, loss G: -0.4915\n",
+            "Epoch [2/10] Batch 3600/3750                   Loss D: -0.0043, loss G: -0.5019\n",
+            "Epoch [2/10] Batch 3700/3750                   Loss D: -0.0027, loss G: -0.5062\n",
+            "Epoch [3/10] Batch 100/3750                   Loss D: -0.0027, loss G: -0.4981\n",
+            "Epoch [3/10] Batch 200/3750                   Loss D: -0.0029, loss G: -0.4924\n",
+            "Epoch [3/10] Batch 300/3750                   Loss D: -0.0026, loss G: -0.5082\n",
+            "Epoch [3/10] Batch 400/3750                   Loss D: -0.0037, loss G: -0.4852\n",
+            "Epoch [3/10] Batch 500/3750                   Loss D: 0.0004, loss G: -0.5082\n",
+            "Epoch [3/10] Batch 600/3750                   Loss D: -0.0006, loss G: -0.4986\n",
+            "Epoch [3/10] Batch 700/3750                   Loss D: -0.0051, loss G: -0.4893\n",
+            "Epoch [3/10] Batch 800/3750                   Loss D: -0.0021, loss G: -0.5034\n",
+            "Epoch [3/10] Batch 900/3750                   Loss D: -0.0001, loss G: -0.5017\n",
+            "Epoch [3/10] Batch 1000/3750                   Loss D: -0.0016, loss G: -0.4961\n",
+            "Epoch [3/10] Batch 1100/3750                   Loss D: -0.0022, loss G: -0.5031\n",
+            "Epoch [3/10] Batch 1200/3750                   Loss D: -0.0043, loss G: -0.5001\n",
+            "Epoch [3/10] Batch 1300/3750                   Loss D: -0.0009, loss G: -0.5042\n",
+            "Epoch [3/10] Batch 1400/3750                   Loss D: -0.0039, loss G: -0.4915\n",
+            "Epoch [3/10] Batch 1500/3750                   Loss D: -0.0063, loss G: -0.5021\n",
+            "Epoch [3/10] Batch 1600/3750                   Loss D: -0.0030, loss G: -0.4951\n",
+            "Epoch [3/10] Batch 1700/3750                   Loss D: -0.0026, loss G: -0.5071\n",
+            "Epoch [3/10] Batch 1800/3750                   Loss D: -0.0024, loss G: -0.4934\n",
+            "Epoch [3/10] Batch 1900/3750                   Loss D: -0.0051, loss G: -0.4925\n",
+            "Epoch [3/10] Batch 2000/3750                   Loss D: -0.0090, loss G: -0.5091\n",
+            "Epoch [3/10] Batch 2100/3750                   Loss D: -0.0024, loss G: -0.4979\n",
+            "Epoch [3/10] Batch 2200/3750                   Loss D: -0.0030, loss G: -0.4936\n",
+            "Epoch [3/10] Batch 2300/3750                   Loss D: -0.0053, loss G: -0.5066\n",
+            "Epoch [3/10] Batch 2400/3750                   Loss D: -0.0047, loss G: -0.4986\n",
+            "Epoch [3/10] Batch 2500/3750                   Loss D: -0.0043, loss G: -0.4995\n",
+            "Epoch [3/10] Batch 2600/3750                   Loss D: -0.0049, loss G: -0.4927\n",
+            "Epoch [3/10] Batch 2700/3750                   Loss D: -0.0065, loss G: -0.4936\n",
+            "Epoch [3/10] Batch 2800/3750                   Loss D: -0.0072, loss G: -0.5075\n",
+            "Epoch [3/10] Batch 2900/3750                   Loss D: -0.0047, loss G: -0.5044\n",
+            "Epoch [3/10] Batch 3000/3750                   Loss D: -0.0041, loss G: -0.4885\n",
+            "Epoch [3/10] Batch 3100/3750                   Loss D: -0.0049, loss G: -0.5111\n",
+            "Epoch [3/10] Batch 3200/3750                   Loss D: -0.0010, loss G: -0.4951\n",
+            "Epoch [3/10] Batch 3300/3750                   Loss D: -0.0062, loss G: -0.5155\n",
+            "Epoch [3/10] Batch 3400/3750                   Loss D: -0.0024, loss G: -0.4925\n",
+            "Epoch [3/10] Batch 3500/3750                   Loss D: -0.0059, loss G: -0.5156\n",
+            "Epoch [3/10] Batch 3600/3750                   Loss D: -0.0030, loss G: -0.4933\n",
+            "Epoch [3/10] Batch 3700/3750                   Loss D: -0.0036, loss G: -0.4990\n",
+            "Epoch [4/10] Batch 100/3750                   Loss D: -0.0026, loss G: -0.4958\n",
+            "Epoch [4/10] Batch 200/3750                   Loss D: -0.0009, loss G: -0.5010\n",
+            "Epoch [4/10] Batch 300/3750                   Loss D: -0.0024, loss G: -0.5063\n",
+            "Epoch [4/10] Batch 400/3750                   Loss D: -0.0052, loss G: -0.5031\n",
+            "Epoch [4/10] Batch 500/3750                   Loss D: -0.0029, loss G: -0.5017\n",
+            "Epoch [4/10] Batch 600/3750                   Loss D: -0.0027, loss G: -0.4863\n",
+            "Epoch [4/10] Batch 700/3750                   Loss D: -0.0037, loss G: -0.4967\n",
+            "Epoch [4/10] Batch 800/3750                   Loss D: -0.0010, loss G: -0.5009\n",
+            "Epoch [4/10] Batch 900/3750                   Loss D: -0.0046, loss G: -0.5101\n",
+            "Epoch [4/10] Batch 1000/3750                   Loss D: -0.0026, loss G: -0.4984\n",
+            "Epoch [4/10] Batch 1100/3750                   Loss D: 0.0007, loss G: -0.4998\n",
+            "Epoch [4/10] Batch 1200/3750                   Loss D: -0.0029, loss G: -0.5035\n",
+            "Epoch [4/10] Batch 1300/3750                   Loss D: -0.0044, loss G: -0.4957\n",
+            "Epoch [4/10] Batch 1400/3750                   Loss D: -0.0033, loss G: -0.5013\n",
+            "Epoch [4/10] Batch 1500/3750                   Loss D: -0.0019, loss G: -0.4905\n",
+            "Epoch [4/10] Batch 1600/3750                   Loss D: -0.0037, loss G: -0.5051\n",
+            "Epoch [4/10] Batch 1700/3750                   Loss D: -0.0051, loss G: -0.4912\n",
+            "Epoch [4/10] Batch 1800/3750                   Loss D: -0.0019, loss G: -0.4996\n",
+            "Epoch [4/10] Batch 1900/3750                   Loss D: -0.0047, loss G: -0.5016\n",
+            "Epoch [4/10] Batch 2000/3750                   Loss D: -0.0039, loss G: -0.4955\n",
+            "Epoch [4/10] Batch 2100/3750                   Loss D: -0.0056, loss G: -0.5072\n",
+            "Epoch [4/10] Batch 2200/3750                   Loss D: -0.0048, loss G: -0.4935\n",
+            "Epoch [4/10] Batch 2300/3750                   Loss D: -0.0035, loss G: -0.5056\n",
+            "Epoch [4/10] Batch 2400/3750                   Loss D: -0.0022, loss G: -0.5014\n",
+            "Epoch [4/10] Batch 2500/3750                   Loss D: -0.0005, loss G: -0.4924\n",
+            "Epoch [4/10] Batch 2600/3750                   Loss D: -0.0054, loss G: -0.5005\n",
+            "Epoch [4/10] Batch 2700/3750                   Loss D: -0.0021, loss G: -0.4993\n",
+            "Epoch [4/10] Batch 2800/3750                   Loss D: -0.0054, loss G: -0.4990\n",
+            "Epoch [4/10] Batch 2900/3750                   Loss D: -0.0009, loss G: -0.5074\n",
+            "Epoch [4/10] Batch 3000/3750                   Loss D: -0.0045, loss G: -0.4914\n",
+            "Epoch [4/10] Batch 3100/3750                   Loss D: -0.0059, loss G: -0.4960\n",
+            "Epoch [4/10] Batch 3200/3750                   Loss D: -0.0033, loss G: -0.5098\n",
+            "Epoch [4/10] Batch 3300/3750                   Loss D: -0.0039, loss G: -0.4936\n",
+            "Epoch [4/10] Batch 3400/3750                   Loss D: -0.0023, loss G: -0.4972\n",
+            "Epoch [4/10] Batch 3500/3750                   Loss D: -0.0019, loss G: -0.5032\n",
+            "Epoch [4/10] Batch 3600/3750                   Loss D: -0.0049, loss G: -0.4915\n",
+            "Epoch [4/10] Batch 3700/3750                   Loss D: -0.0022, loss G: -0.5065\n",
+            "Epoch [5/10] Batch 100/3750                   Loss D: -0.0025, loss G: -0.4994\n",
+            "Epoch [5/10] Batch 200/3750                   Loss D: -0.0034, loss G: -0.5005\n",
+            "Epoch [5/10] Batch 300/3750                   Loss D: -0.0050, loss G: -0.5015\n",
+            "Epoch [5/10] Batch 400/3750                   Loss D: -0.0036, loss G: -0.4977\n",
+            "Epoch [5/10] Batch 500/3750                   Loss D: -0.0033, loss G: -0.4982\n",
+            "Epoch [5/10] Batch 600/3750                   Loss D: -0.0044, loss G: -0.4963\n",
+            "Epoch [5/10] Batch 700/3750                   Loss D: -0.0036, loss G: -0.5141\n",
+            "Epoch [5/10] Batch 800/3750                   Loss D: -0.0025, loss G: -0.4914\n",
+            "Epoch [5/10] Batch 900/3750                   Loss D: -0.0025, loss G: -0.4961\n",
+            "Epoch [5/10] Batch 1000/3750                   Loss D: -0.0040, loss G: -0.4935\n",
+            "Epoch [5/10] Batch 1100/3750                   Loss D: -0.0028, loss G: -0.4985\n",
+            "Epoch [5/10] Batch 1200/3750                   Loss D: -0.0052, loss G: -0.4888\n",
+            "Epoch [5/10] Batch 1300/3750                   Loss D: -0.0043, loss G: -0.4984\n",
+            "Epoch [5/10] Batch 1400/3750                   Loss D: -0.0039, loss G: -0.4922\n",
+            "Epoch [5/10] Batch 1500/3750                   Loss D: -0.0033, loss G: -0.4976\n",
+            "Epoch [5/10] Batch 1600/3750                   Loss D: -0.0014, loss G: -0.4988\n",
+            "Epoch [5/10] Batch 1700/3750                   Loss D: -0.0069, loss G: -0.5106\n",
+            "Epoch [5/10] Batch 1800/3750                   Loss D: 0.0003, loss G: -0.4913\n",
+            "Epoch [5/10] Batch 1900/3750                   Loss D: -0.0041, loss G: -0.4869\n",
+            "Epoch [5/10] Batch 2000/3750                   Loss D: -0.0028, loss G: -0.5061\n",
+            "Epoch [5/10] Batch 2100/3750                   Loss D: -0.0032, loss G: -0.4949\n",
+            "Epoch [5/10] Batch 2200/3750                   Loss D: -0.0006, loss G: -0.4980\n",
+            "Epoch [5/10] Batch 2300/3750                   Loss D: -0.0043, loss G: -0.5005\n",
+            "Epoch [5/10] Batch 2400/3750                   Loss D: -0.0028, loss G: -0.5026\n",
+            "Epoch [5/10] Batch 2500/3750                   Loss D: -0.0015, loss G: -0.4894\n",
+            "Epoch [5/10] Batch 2600/3750                   Loss D: -0.0016, loss G: -0.5002\n",
+            "Epoch [5/10] Batch 2700/3750                   Loss D: -0.0027, loss G: -0.5040\n",
+            "Epoch [5/10] Batch 2800/3750                   Loss D: -0.0024, loss G: -0.4950\n",
+            "Epoch [5/10] Batch 2900/3750                   Loss D: -0.0043, loss G: -0.5065\n",
+            "Epoch [5/10] Batch 3000/3750                   Loss D: -0.0023, loss G: -0.4973\n",
+            "Epoch [5/10] Batch 3100/3750                   Loss D: -0.0054, loss G: -0.5013\n",
+            "Epoch [5/10] Batch 3200/3750                   Loss D: -0.0055, loss G: -0.4915\n",
+            "Epoch [5/10] Batch 3300/3750                   Loss D: -0.0005, loss G: -0.5003\n",
+            "Epoch [5/10] Batch 3400/3750                   Loss D: 0.0021, loss G: -0.5123\n",
+            "Epoch [5/10] Batch 3500/3750                   Loss D: -0.0025, loss G: -0.5061\n",
+            "Epoch [5/10] Batch 3600/3750                   Loss D: -0.0056, loss G: -0.4878\n",
+            "Epoch [5/10] Batch 3700/3750                   Loss D: -0.0010, loss G: -0.4904\n",
+            "Epoch [6/10] Batch 100/3750                   Loss D: -0.0011, loss G: -0.5075\n",
+            "Epoch [6/10] Batch 200/3750                   Loss D: -0.0012, loss G: -0.4982\n",
+            "Epoch [6/10] Batch 300/3750                   Loss D: -0.0027, loss G: -0.4844\n",
+            "Epoch [6/10] Batch 400/3750                   Loss D: -0.0009, loss G: -0.5052\n",
+            "Epoch [6/10] Batch 500/3750                   Loss D: -0.0051, loss G: -0.5053\n",
+            "Epoch [6/10] Batch 600/3750                   Loss D: -0.0033, loss G: -0.4884\n",
+            "Epoch [6/10] Batch 700/3750                   Loss D: -0.0037, loss G: -0.5062\n",
+            "Epoch [6/10] Batch 800/3750                   Loss D: 0.0000, loss G: -0.5074\n",
+            "Epoch [6/10] Batch 900/3750                   Loss D: -0.0024, loss G: -0.4913\n",
+            "Epoch [6/10] Batch 1000/3750                   Loss D: 0.0030, loss G: -0.4941\n",
+            "Epoch [6/10] Batch 1100/3750                   Loss D: -0.0030, loss G: -0.5037\n",
+            "Epoch [6/10] Batch 1200/3750                   Loss D: -0.0021, loss G: -0.4996\n",
+            "Epoch [6/10] Batch 1300/3750                   Loss D: -0.0017, loss G: -0.4996\n",
+            "Epoch [6/10] Batch 1400/3750                   Loss D: -0.0006, loss G: -0.4933\n",
+            "Epoch [6/10] Batch 1500/3750                   Loss D: 0.0062, loss G: -0.5148\n",
+            "Epoch [6/10] Batch 1600/3750                   Loss D: -0.0035, loss G: -0.4997\n",
+            "Epoch [6/10] Batch 1700/3750                   Loss D: -0.0019, loss G: -0.4888\n",
+            "Epoch [6/10] Batch 1800/3750                   Loss D: -0.0025, loss G: -0.5085\n",
+            "Epoch [6/10] Batch 1900/3750                   Loss D: -0.0057, loss G: -0.5050\n",
+            "Epoch [6/10] Batch 2000/3750                   Loss D: -0.0018, loss G: -0.4961\n",
+            "Epoch [6/10] Batch 2100/3750                   Loss D: -0.0034, loss G: -0.4926\n",
+            "Epoch [6/10] Batch 2200/3750                   Loss D: -0.0040, loss G: -0.4948\n",
+            "Epoch [6/10] Batch 2300/3750                   Loss D: 0.0001, loss G: -0.5038\n",
+            "Epoch [6/10] Batch 2400/3750                   Loss D: 0.0039, loss G: -0.5102\n",
+            "Epoch [6/10] Batch 2500/3750                   Loss D: 0.0011, loss G: -0.4952\n",
+            "Epoch [6/10] Batch 2600/3750                   Loss D: -0.0023, loss G: -0.4910\n",
+            "Epoch [6/10] Batch 2700/3750                   Loss D: -0.0027, loss G: -0.5096\n",
+            "Epoch [6/10] Batch 2800/3750                   Loss D: -0.0070, loss G: -0.5112\n",
+            "Epoch [6/10] Batch 2900/3750                   Loss D: -0.0019, loss G: -0.4948\n",
+            "Epoch [6/10] Batch 3000/3750                   Loss D: -0.0026, loss G: -0.4915\n",
+            "Epoch [6/10] Batch 3100/3750                   Loss D: -0.0029, loss G: -0.4974\n",
+            "Epoch [6/10] Batch 3200/3750                   Loss D: -0.0053, loss G: -0.5091\n",
+            "Epoch [6/10] Batch 3300/3750                   Loss D: -0.0027, loss G: -0.4999\n",
+            "Epoch [6/10] Batch 3400/3750                   Loss D: -0.0028, loss G: -0.4978\n",
+            "Epoch [6/10] Batch 3500/3750                   Loss D: -0.0019, loss G: -0.5035\n",
+            "Epoch [6/10] Batch 3600/3750                   Loss D: -0.0014, loss G: -0.4923\n",
+            "Epoch [6/10] Batch 3700/3750                   Loss D: -0.0024, loss G: -0.4987\n",
+            "Epoch [7/10] Batch 100/3750                   Loss D: -0.0047, loss G: -0.5067\n",
+            "Epoch [7/10] Batch 200/3750                   Loss D: -0.0064, loss G: -0.4877\n",
+            "Epoch [7/10] Batch 300/3750                   Loss D: -0.0028, loss G: -0.5004\n",
+            "Epoch [7/10] Batch 400/3750                   Loss D: -0.0040, loss G: -0.4992\n",
+            "Epoch [7/10] Batch 500/3750                   Loss D: -0.0020, loss G: -0.5005\n",
+            "Epoch [7/10] Batch 600/3750                   Loss D: -0.0026, loss G: -0.5004\n",
+            "Epoch [7/10] Batch 700/3750                   Loss D: -0.0010, loss G: -0.5040\n",
+            "Epoch [7/10] Batch 800/3750                   Loss D: -0.0050, loss G: -0.4982\n",
+            "Epoch [7/10] Batch 900/3750                   Loss D: -0.0040, loss G: -0.4954\n",
+            "Epoch [7/10] Batch 1000/3750                   Loss D: -0.0051, loss G: -0.4927\n",
+            "Epoch [7/10] Batch 1100/3750                   Loss D: -0.0013, loss G: -0.4957\n",
+            "Epoch [7/10] Batch 1200/3750                   Loss D: 0.0025, loss G: -0.5183\n",
+            "Epoch [7/10] Batch 1300/3750                   Loss D: -0.0030, loss G: -0.4889\n",
+            "Epoch [7/10] Batch 1400/3750                   Loss D: -0.0030, loss G: -0.4885\n",
+            "Epoch [7/10] Batch 1500/3750                   Loss D: -0.0034, loss G: -0.5052\n",
+            "Epoch [7/10] Batch 1600/3750                   Loss D: -0.0031, loss G: -0.4985\n",
+            "Epoch [7/10] Batch 1700/3750                   Loss D: -0.0030, loss G: -0.4966\n",
+            "Epoch [7/10] Batch 1800/3750                   Loss D: -0.0015, loss G: -0.5048\n",
+            "Epoch [7/10] Batch 1900/3750                   Loss D: -0.0026, loss G: -0.4986\n",
+            "Epoch [7/10] Batch 2000/3750                   Loss D: -0.0030, loss G: -0.5044\n",
+            "Epoch [7/10] Batch 2100/3750                   Loss D: -0.0025, loss G: -0.4895\n",
+            "Epoch [7/10] Batch 2200/3750                   Loss D: -0.0022, loss G: -0.4996\n",
+            "Epoch [7/10] Batch 2300/3750                   Loss D: -0.0007, loss G: -0.5045\n",
+            "Epoch [7/10] Batch 2400/3750                   Loss D: -0.0024, loss G: -0.4849\n",
+            "Epoch [7/10] Batch 2500/3750                   Loss D: -0.0053, loss G: -0.5045\n",
+            "Epoch [7/10] Batch 2600/3750                   Loss D: -0.0017, loss G: -0.4959\n",
+            "Epoch [7/10] Batch 2700/3750                   Loss D: -0.0008, loss G: -0.4996\n",
+            "Epoch [7/10] Batch 2800/3750                   Loss D: -0.0026, loss G: -0.5021\n",
+            "Epoch [7/10] Batch 2900/3750                   Loss D: -0.0024, loss G: -0.4979\n",
+            "Epoch [7/10] Batch 3000/3750                   Loss D: -0.0027, loss G: -0.4986\n",
+            "Epoch [7/10] Batch 3100/3750                   Loss D: -0.0033, loss G: -0.5038\n",
+            "Epoch [7/10] Batch 3200/3750                   Loss D: -0.0062, loss G: -0.4876\n",
+            "Epoch [7/10] Batch 3300/3750                   Loss D: -0.0021, loss G: -0.4927\n",
+            "Epoch [7/10] Batch 3400/3750                   Loss D: -0.0034, loss G: -0.5103\n",
+            "Epoch [7/10] Batch 3500/3750                   Loss D: -0.0023, loss G: -0.4956\n",
+            "Epoch [7/10] Batch 3600/3750                   Loss D: -0.0041, loss G: -0.4939\n",
+            "Epoch [7/10] Batch 3700/3750                   Loss D: -0.0046, loss G: -0.5083\n",
+            "Epoch [8/10] Batch 100/3750                   Loss D: -0.0029, loss G: -0.4985\n",
+            "Epoch [8/10] Batch 200/3750                   Loss D: -0.0009, loss G: -0.5109\n",
+            "Epoch [8/10] Batch 300/3750                   Loss D: -0.0048, loss G: -0.4945\n",
+            "Epoch [8/10] Batch 400/3750                   Loss D: -0.0041, loss G: -0.4953\n",
+            "Epoch [8/10] Batch 500/3750                   Loss D: -0.0009, loss G: -0.4949\n",
+            "Epoch [8/10] Batch 600/3750                   Loss D: -0.0021, loss G: -0.5110\n",
+            "Epoch [8/10] Batch 700/3750                   Loss D: -0.0038, loss G: -0.4874\n",
+            "Epoch [8/10] Batch 800/3750                   Loss D: -0.0010, loss G: -0.4979\n",
+            "Epoch [8/10] Batch 900/3750                   Loss D: -0.0024, loss G: -0.5055\n",
+            "Epoch [8/10] Batch 1000/3750                   Loss D: -0.0019, loss G: -0.4990\n",
+            "Epoch [8/10] Batch 1100/3750                   Loss D: -0.0059, loss G: -0.4890\n",
+            "Epoch [8/10] Batch 1200/3750                   Loss D: -0.0031, loss G: -0.5019\n",
+            "Epoch [8/10] Batch 1300/3750                   Loss D: -0.0001, loss G: -0.5003\n",
+            "Epoch [8/10] Batch 1400/3750                   Loss D: -0.0022, loss G: -0.4961\n",
+            "Epoch [8/10] Batch 1500/3750                   Loss D: -0.0039, loss G: -0.4850\n",
+            "Epoch [8/10] Batch 1600/3750                   Loss D: -0.0001, loss G: -0.4995\n",
+            "Epoch [8/10] Batch 1700/3750                   Loss D: -0.0017, loss G: -0.4933\n",
+            "Epoch [8/10] Batch 1800/3750                   Loss D: -0.0023, loss G: -0.4929\n",
+            "Epoch [8/10] Batch 1900/3750                   Loss D: -0.0082, loss G: -0.5114\n",
+            "Epoch [8/10] Batch 2000/3750                   Loss D: -0.0029, loss G: -0.4958\n",
+            "Epoch [8/10] Batch 2100/3750                   Loss D: -0.0037, loss G: -0.4947\n",
+            "Epoch [8/10] Batch 2200/3750                   Loss D: -0.0033, loss G: -0.4997\n",
+            "Epoch [8/10] Batch 2300/3750                   Loss D: -0.0022, loss G: -0.5159\n",
+            "Epoch [8/10] Batch 2400/3750                   Loss D: -0.0018, loss G: -0.4965\n",
+            "Epoch [8/10] Batch 2500/3750                   Loss D: -0.0051, loss G: -0.4886\n",
+            "Epoch [8/10] Batch 2600/3750                   Loss D: -0.0039, loss G: -0.5092\n",
+            "Epoch [8/10] Batch 2700/3750                   Loss D: -0.0042, loss G: -0.4885\n",
+            "Epoch [8/10] Batch 2800/3750                   Loss D: -0.0001, loss G: -0.4892\n",
+            "Epoch [8/10] Batch 2900/3750                   Loss D: -0.0054, loss G: -0.5047\n",
+            "Epoch [8/10] Batch 3000/3750                   Loss D: -0.0019, loss G: -0.4976\n",
+            "Epoch [8/10] Batch 3100/3750                   Loss D: -0.0020, loss G: -0.4906\n",
+            "Epoch [8/10] Batch 3200/3750                   Loss D: -0.0027, loss G: -0.4951\n",
+            "Epoch [8/10] Batch 3300/3750                   Loss D: -0.0041, loss G: -0.4963\n",
+            "Epoch [8/10] Batch 3400/3750                   Loss D: -0.0067, loss G: -0.5086\n",
+            "Epoch [8/10] Batch 3500/3750                   Loss D: -0.0049, loss G: -0.4841\n",
+            "Epoch [8/10] Batch 3600/3750                   Loss D: -0.0020, loss G: -0.5020\n",
+            "Epoch [8/10] Batch 3700/3750                   Loss D: -0.0010, loss G: -0.4989\n",
+            "Epoch [9/10] Batch 100/3750                   Loss D: -0.0031, loss G: -0.4877\n",
+            "Epoch [9/10] Batch 200/3750                   Loss D: -0.0038, loss G: -0.5043\n",
+            "Epoch [9/10] Batch 300/3750                   Loss D: -0.0012, loss G: -0.5083\n",
+            "Epoch [9/10] Batch 400/3750                   Loss D: -0.0023, loss G: -0.4994\n",
+            "Epoch [9/10] Batch 500/3750                   Loss D: -0.0025, loss G: -0.4888\n",
+            "Epoch [9/10] Batch 600/3750                   Loss D: -0.0004, loss G: -0.5074\n",
+            "Epoch [9/10] Batch 700/3750                   Loss D: -0.0021, loss G: -0.5071\n",
+            "Epoch [9/10] Batch 800/3750                   Loss D: -0.0026, loss G: -0.4994\n",
+            "Epoch [9/10] Batch 900/3750                   Loss D: -0.0026, loss G: -0.4893\n",
+            "Epoch [9/10] Batch 1000/3750                   Loss D: -0.0021, loss G: -0.4991\n",
+            "Epoch [9/10] Batch 1100/3750                   Loss D: 0.0005, loss G: -0.5113\n",
+            "Epoch [9/10] Batch 1200/3750                   Loss D: -0.0019, loss G: -0.4907\n",
+            "Epoch [9/10] Batch 1300/3750                   Loss D: -0.0030, loss G: -0.4914\n",
+            "Epoch [9/10] Batch 1400/3750                   Loss D: -0.0027, loss G: -0.5031\n",
+            "Epoch [9/10] Batch 1500/3750                   Loss D: -0.0029, loss G: -0.5028\n",
+            "Epoch [9/10] Batch 1600/3750                   Loss D: -0.0011, loss G: -0.4989\n",
+            "Epoch [9/10] Batch 1700/3750                   Loss D: -0.0051, loss G: -0.5052\n",
+            "Epoch [9/10] Batch 1800/3750                   Loss D: -0.0021, loss G: -0.4996\n",
+            "Epoch [9/10] Batch 1900/3750                   Loss D: -0.0019, loss G: -0.5016\n",
+            "Epoch [9/10] Batch 2000/3750                   Loss D: -0.0046, loss G: -0.4891\n",
+            "Epoch [9/10] Batch 2100/3750                   Loss D: -0.0041, loss G: -0.4936\n",
+            "Epoch [9/10] Batch 2200/3750                   Loss D: -0.0056, loss G: -0.5113\n",
+            "Epoch [9/10] Batch 2300/3750                   Loss D: -0.0013, loss G: -0.4882\n",
+            "Epoch [9/10] Batch 2400/3750                   Loss D: 0.0018, loss G: -0.4893\n",
+            "Epoch [9/10] Batch 2500/3750                   Loss D: -0.0021, loss G: -0.5094\n",
+            "Epoch [9/10] Batch 2600/3750                   Loss D: -0.0040, loss G: -0.5018\n",
+            "Epoch [9/10] Batch 2700/3750                   Loss D: -0.0017, loss G: -0.4905\n",
+            "Epoch [9/10] Batch 2800/3750                   Loss D: -0.0016, loss G: -0.5050\n",
+            "Epoch [9/10] Batch 2900/3750                   Loss D: -0.0025, loss G: -0.4918\n",
+            "Epoch [9/10] Batch 3000/3750                   Loss D: -0.0036, loss G: -0.4962\n",
+            "Epoch [9/10] Batch 3100/3750                   Loss D: -0.0008, loss G: -0.5071\n",
+            "Epoch [9/10] Batch 3200/3750                   Loss D: -0.0034, loss G: -0.4921\n",
+            "Epoch [9/10] Batch 3300/3750                   Loss D: -0.0017, loss G: -0.5013\n",
+            "Epoch [9/10] Batch 3400/3750                   Loss D: -0.0035, loss G: -0.5001\n",
+            "Epoch [9/10] Batch 3500/3750                   Loss D: -0.0003, loss G: -0.5042\n",
+            "Epoch [9/10] Batch 3600/3750                   Loss D: -0.0024, loss G: -0.4967\n",
+            "Epoch [9/10] Batch 3700/3750                   Loss D: -0.0040, loss G: -0.5010\n",
+            "Epoch [10/10] Batch 100/3750                   Loss D: -0.0025, loss G: -0.4959\n",
+            "Epoch [10/10] Batch 200/3750                   Loss D: -0.0025, loss G: -0.4948\n",
+            "Epoch [10/10] Batch 300/3750                   Loss D: -0.0028, loss G: -0.5041\n",
+            "Epoch [10/10] Batch 400/3750                   Loss D: -0.0043, loss G: -0.4954\n",
+            "Epoch [10/10] Batch 500/3750                   Loss D: -0.0039, loss G: -0.4957\n",
+            "Epoch [10/10] Batch 600/3750                   Loss D: -0.0030, loss G: -0.5023\n",
+            "Epoch [10/10] Batch 700/3750                   Loss D: -0.0014, loss G: -0.4961\n",
+            "Epoch [10/10] Batch 800/3750                   Loss D: -0.0025, loss G: -0.5009\n",
+            "Epoch [10/10] Batch 900/3750                   Loss D: -0.0005, loss G: -0.4962\n",
+            "Epoch [10/10] Batch 1000/3750                   Loss D: -0.0012, loss G: -0.5014\n",
+            "Epoch [10/10] Batch 1100/3750                   Loss D: -0.0035, loss G: -0.4973\n",
+            "Epoch [10/10] Batch 1200/3750                   Loss D: -0.0008, loss G: -0.4894\n",
+            "Epoch [10/10] Batch 1300/3750                   Loss D: -0.0031, loss G: -0.5034\n",
+            "Epoch [10/10] Batch 1400/3750                   Loss D: -0.0001, loss G: -0.5063\n",
+            "Epoch [10/10] Batch 1500/3750                   Loss D: -0.0028, loss G: -0.4955\n",
+            "Epoch [10/10] Batch 1600/3750                   Loss D: -0.0026, loss G: -0.4992\n",
+            "Epoch [10/10] Batch 1700/3750                   Loss D: -0.0031, loss G: -0.5024\n",
+            "Epoch [10/10] Batch 1800/3750                   Loss D: -0.0029, loss G: -0.4947\n",
+            "Epoch [10/10] Batch 1900/3750                   Loss D: 0.0011, loss G: -0.4962\n",
+            "Epoch [10/10] Batch 2000/3750                   Loss D: -0.0056, loss G: -0.5062\n",
+            "Epoch [10/10] Batch 2100/3750                   Loss D: -0.0021, loss G: -0.4998\n",
+            "Epoch [10/10] Batch 2200/3750                   Loss D: -0.0057, loss G: -0.4840\n",
+            "Epoch [10/10] Batch 2300/3750                   Loss D: -0.0029, loss G: -0.4959\n",
+            "Epoch [10/10] Batch 2400/3750                   Loss D: -0.0036, loss G: -0.5013\n",
+            "Epoch [10/10] Batch 2500/3750                   Loss D: -0.0038, loss G: -0.4980\n",
+            "Epoch [10/10] Batch 2600/3750                   Loss D: -0.0046, loss G: -0.4966\n",
+            "Epoch [10/10] Batch 2700/3750                   Loss D: -0.0030, loss G: -0.5035\n",
+            "Epoch [10/10] Batch 2800/3750                   Loss D: -0.0024, loss G: -0.4980\n",
+            "Epoch [10/10] Batch 2900/3750                   Loss D: -0.0032, loss G: -0.4970\n",
+            "Epoch [10/10] Batch 3000/3750                   Loss D: -0.0036, loss G: -0.4967\n",
+            "Epoch [10/10] Batch 3100/3750                   Loss D: -0.0034, loss G: -0.4930\n",
+            "Epoch [10/10] Batch 3200/3750                   Loss D: -0.0037, loss G: -0.5034\n",
+            "Epoch [10/10] Batch 3300/3750                   Loss D: -0.0041, loss G: -0.4950\n",
+            "Epoch [10/10] Batch 3400/3750                   Loss D: -0.0041, loss G: -0.5000\n",
+            "Epoch [10/10] Batch 3500/3750                   Loss D: -0.0026, loss G: -0.4896\n",
+            "Epoch [10/10] Batch 3600/3750                   Loss D: -0.0034, loss G: -0.4970\n",
+            "Epoch [10/10] Batch 3700/3750                   Loss D: -0.0031, loss G: -0.4990\n"
+          ]
+        }
+      ],
+      "source": [
+        "ERRD_x = np.zeros(config.num_epochs)\n",
+        "ERRD_z = np.zeros(config.num_epochs)\n",
+        "ERRG = np.zeros(config.num_epochs)\n",
+        "N = len(dataloader)\n",
+        "#WGAN modified of DCGAN in:\n",
+        "#1. remove sigmoid in the last layer of discriminator(classification -> regression)\n",
+        "#2. no log Loss (Wasserstein distance)\n",
+        "#3. clip param norm to c (Wasserstein distance and Lipschitz continuity)\n",
+        "#4. No momentum-based optimizer, use RMSProp，SGD instead\n",
+        "\n",
+        "for epoch in range(config.num_epochs):\n",
+        "    for iteration, (images, cat) in enumerate(dataloader):\n",
+        "        # Train Critic: max E[critic(real)] - E[critic(fake)]\n",
+        "        # CRITIC_ITERATIONS = 5\n",
+        "        # optim_D.zero_grad()\n",
+        "        ####### \n",
+        "        # Discriminator stage: maximize log(D(x)) + log(1 - D(G(z))) \n",
+        "        #######\n",
+        "        discriminator.zero_grad()\n",
+        "        # real\n",
+        "        label.data.fill_(real_label)\n",
+        "        input_data = images.view(images.shape[0], -1)\n",
+        "        output_real = discriminator(input_data)\n",
+        "        # errD_x = criterion(output, label)\n",
+        "        # ERRD_x[epoch] += errD_x.item()\n",
+        "        # errD_x.backward()\n",
+        "        \n",
+        "        # fake \n",
+        "        noise.data.normal_(0, 1)\n",
+        "        fake = generator(noise)\n",
+        "        label.data.fill_(fake_label)\n",
+        "        output_fake = discriminator(fake.detach())\n",
+        "        # errD_z = criterion(output, label)\n",
+        "        # ERRD_z[epoch] += errD_z.item()\n",
+        "        # errD_z.backward()\n",
+        "        loss_critic = -(torch.mean(output_real) - torch.mean(output_fake))\n",
+        "\n",
+        "\n",
+        "        loss_critic.backward(retain_graph=True)\n",
+        "        optim_D.step()\n",
+        "        \n",
+        "        # modification3: clip param norm to c=0.01 (Wasserstein distance and Lipschitz continuity)\n",
+        "        for parm in discriminator.parameters():\n",
+        "            parm.data.clamp_(-0.01, 0.01)\n",
+        "\n",
+        "\n",
+        "        # Train Generator: max E[critic(gen_fake)] <-> min -E[critic(gen_fake)]\n",
+        "        generator.zero_grad()\n",
+        "        label.data.fill_(real_label)\n",
+        "        gen_fake = discriminator(fake)\n",
+        "        loss_gen = -torch.mean(gen_fake)\n",
+        "\n",
+        "        loss_gen.backward()\n",
+        "\n",
+        "        optim_G.step()\n",
+        "\n",
+        "        ####### \n",
+        "        # Generator stage: maximize log(D(G(x))\n",
+        "        #######\n",
+        "        # generator.zero_grad()\n",
+        "        # label.data.fill_(real_label)\n",
+        "        # output = discriminator(fake)\n",
+        "        # errG = criterion(output, label)\n",
+        "        # ERRG[epoch] += errG.item()\n",
+        "        # errG.backward()\n",
+        "        \n",
+        "        # optim_G.step()\n",
+        "        \n",
+        "        '''\n",
+        "        if (iteration+1) % config.print_freq == 0:\n",
+        "            print('Epoch:{} Iter: {} errD_x: {:.2f} errD_z: {:.2f} errG: {:.2f}'.format(epoch+1,\n",
+        "                                                                                            iteration+1, \n",
+        "        '''\n",
+        "        \n",
+        "        # Print losses occasionally and print to tensorboard\n",
+        "        if iteration % 100 == 0 and iteration > 0:\n",
+        "            generator.eval()\n",
+        "            discriminator.eval()\n",
+        "            print(\n",
+        "                f\"Epoch [{epoch+1}/{10}] Batch {iteration}/{len(dataloader)} \\\n",
+        "                  Loss D: {loss_critic:.4f}, loss G: {loss_gen:.4f}\"\n",
+        "            )\n",
+        "\n",
+        "            with torch.no_grad():\n",
+        "                fake = generator(noise)\n",
+        "                # take out (up to) 32 examples\n",
+        "                img_grid_real = torchvision.utils.make_grid(\n",
+        "                    images[:32], normalize=True\n",
+        "                )\n",
+        "                img_grid_fake = torchvision.utils.make_grid(\n",
+        "                    fake[:32], normalize=True\n",
+        "                )\n",
+        "\n",
+        "                writer_real.add_image(\"Real\", img_grid_real, global_step=step)\n",
+        "                writer_fake.add_image(\"Fake\", img_grid_fake, global_step=step)\n",
+        "\n",
+        "            step += 1\n",
+        "            generator.train()\n",
+        "            discriminator.train()\n"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 13,
+      "metadata": {
+        "id": "sDYUxxGIUdru",
+        "outputId": "10c9d0f3-f820-45bb-f763-3c9f0b84f136",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 401
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 432x504 with 16 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {
+            "needs_background": "light"
+          }
+        }
+      ],
+      "source": [
+        "noise.data.normal_(0, 1)\n",
+        "fake = generator(noise)\n",
+        "\n",
+        "plt.figure(figsize=(6, 7))\n",
+        "for i in range(16):\n",
+        "    plt.subplot(4, 4, i + 1)\n",
+        "    plt.imshow(fake[i].detach().numpy().reshape(28, 28), cmap=plt.cm.Greys_r)\n",
+        "    plt.axis('off')"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "def gradient_penalty(critic, real, fake):\n",
+        "    real = real.view(16, 784)\n",
+        "    alpha = torch.rand((16, 784))\n",
+        "    interpolated_images = real * alpha + fake * (1 - alpha)\n",
+        "\n",
+        "    # Calculate critic scores\n",
+        "    mixed_scores = critic(interpolated_images)\n",
+        "\n",
+        "    # Take the gradient of the scores with respect to the images\n",
+        "    gradient = torch.autograd.grad(\n",
+        "        inputs=interpolated_images,\n",
+        "        outputs=mixed_scores,\n",
+        "        grad_outputs=torch.ones_like(mixed_scores),\n",
+        "        create_graph=True,\n",
+        "        retain_graph=True,\n",
+        "    )[0]\n",
+        "    gradient = gradient.view(gradient.shape[0], -1)\n",
+        "    gradient_norm = gradient.norm(2, dim=1)\n",
+        "    gradient_penalty = torch.mean((gradient_norm - 1) ** 2)\n",
+        "    return gradient_penalty"
+      ],
+      "metadata": {
+        "id": "VxNbDtnNXcPW"
+      },
+      "execution_count": 14,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# switch algorighm back to Adam\n",
+        "optimD = optim.Adam(discriminator.parameters(), lr=1e-4, betas=(0.5, 0.9))\n",
+        "optimG = optim.Adam(generator.parameters(), lr=1e-4, betas=(0.5, 0.9))"
+      ],
+      "metadata": {
+        "id": "uL4U6pBCXfJy"
+      },
+      "execution_count": 15,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "for epoch in range(config.num_epochs):\n",
+        "    for iteration, (images, cat) in enumerate(dataloader):\n",
+        "        # real\n",
+        "        label.data.fill_(real_label)\n",
+        "        input_data = images.view(images.shape[0], -1)\n",
+        "        output_real = discriminator(input_data)\n",
+        "        # errD_x = criterion(output, label)\n",
+        "        # ERRD_x[epoch] += errD_x.item()\n",
+        "        # errD_x.backward()\n",
+        "        \n",
+        "        # fake \n",
+        "        noise.data.normal_(0, 1)\n",
+        "        fake = generator(noise)\n",
+        "        label.data.fill_(fake_label)\n",
+        "        output_fake = discriminator(fake.detach())\n",
+        "        \n",
+        "        # add gp here\n",
+        "        gp = gradient_penalty(discriminator, images, fake)\n",
+        "        loss_critic = (\n",
+        "                -(torch.mean(output_real) - torch.mean(output_fake)) + 10 * gp # LAMBDA_GP = 10\n",
+        "            )\n",
+        "        discriminator.zero_grad()\n",
+        "        loss_critic.backward(retain_graph=True)\n",
+        "        optim_D.step()\n",
+        "\n",
+        "\n",
+        "\n",
+        "        # Train Generator: max E[critic(gen_fake)] <-> min -E[critic(gen_fake)]\n",
+        "        generator.zero_grad()\n",
+        "        label.data.fill_(real_label)\n",
+        "        gen_fake = discriminator(fake)\n",
+        "        loss_gen = -torch.mean(gen_fake)\n",
+        "\n",
+        "        loss_gen.backward()\n",
+        "\n",
+        "        optim_G.step()\n",
+        "        ####### \n",
+        "        # Generator stage: maximize log(D(G(x))\n",
+        "        #######\n",
+        "        # generator.zero_grad()\n",
+        "        # label.data.fill_(real_label)\n",
+        "        # output = discriminator(fake)\n",
+        "        # errG = criterion(output, label)\n",
+        "        # ERRG[epoch] += errG.item()\n",
+        "        # errG.backward()\n",
+        "        \n",
+        "        # optim_G.step()\n",
+        "        '''\n",
+        "        if (iteration+1) % config.print_freq == 0:\n",
+        "            print('Epoch:{} Iter: {} errD_x: {:.2f} errD_z: {:.2f} errG: {:.2f}'.format(epoch+1,\n",
+        "                                                                                            iteration+1, \n",
+        "        '''\n",
+        "        \n",
+        "        # Print losses occasionally and print to tensorboard\n",
+        "        if iteration % 100 == 0 and iteration > 0:\n",
+        "            generator.eval()\n",
+        "            discriminator.eval()\n",
+        "            print(\n",
+        "                f\"Epoch [{epoch+1}/{10}] Batch {iteration}/{len(dataloader)} \\\n",
+        "                  Loss D: {loss_critic:.4f}, loss G: {loss_gen:.4f}\"\n",
+        "            )\n",
+        "\n",
+        "            with torch.no_grad():\n",
+        "                fake = generator(noise)\n",
+        "                # take out (up to) 32 examples\n",
+        "                img_grid_real = torchvision.utils.make_grid(\n",
+        "                    images[:32], normalize=True\n",
+        "                )\n",
+        "                img_grid_fake = torchvision.utils.make_grid(\n",
+        "                    fake[:32], normalize=True\n",
+        "                )\n",
+        "\n",
+        "                writer_real.add_image(\"Real\", img_grid_real, global_step=step)\n",
+        "                writer_fake.add_image(\"Fake\", img_grid_fake, global_step=step)\n",
+        "\n",
+        "            step += 1\n",
+        "            generator.train()\n",
+        "            discriminator.train()"
+      ],
+      "metadata": {
+        "id": "LAbneMp4XifF",
+        "outputId": "cdb96d26-aef3-4598-e0a9-29f3a368f2ac",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "execution_count": 16,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Epoch [1/10] Batch 100/3750                   Loss D: 8.9805, loss G: -0.5894\n",
+            "Epoch [1/10] Batch 200/3750                   Loss D: 9.2363, loss G: -0.5446\n",
+            "Epoch [1/10] Batch 300/3750                   Loss D: 9.8855, loss G: -0.4985\n",
+            "Epoch [1/10] Batch 400/3750                   Loss D: 10.0112, loss G: -0.4965\n",
+            "Epoch [1/10] Batch 500/3750                   Loss D: 10.0004, loss G: -0.4967\n",
+            "Epoch [1/10] Batch 600/3750                   Loss D: 10.0002, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 700/3750                   Loss D: 10.0006, loss G: -0.4968\n",
+            "Epoch [1/10] Batch 800/3750                   Loss D: 10.0011, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4964\n",
+            "Epoch [1/10] Batch 1500/3750                   Loss D: 9.9741, loss G: -0.4985\n",
+            "Epoch [1/10] Batch 1600/3750                   Loss D: 9.8007, loss G: -0.5206\n",
+            "Epoch [1/10] Batch 1700/3750                   Loss D: 9.8936, loss G: -0.5285\n",
+            "Epoch [1/10] Batch 1800/3750                   Loss D: 9.8411, loss G: -0.5217\n",
+            "Epoch [1/10] Batch 1900/3750                   Loss D: 10.0126, loss G: -0.5267\n",
+            "Epoch [1/10] Batch 2000/3750                   Loss D: 10.0017, loss G: -0.4925\n",
+            "Epoch [1/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4909\n",
+            "Epoch [1/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [1/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [1/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [1/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4908\n",
+            "Epoch [2/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4906\n",
+            "Epoch [2/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4906\n",
+            "Epoch [2/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4906\n",
+            "Epoch [2/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4906\n",
+            "Epoch [2/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4905\n",
+            "Epoch [2/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4904\n",
+            "Epoch [2/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [2/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4903\n",
+            "Epoch [3/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [3/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4902\n",
+            "Epoch [4/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [4/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4900\n",
+            "Epoch [5/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [5/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [6/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [7/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [8/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [9/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 1900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2700/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2800/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 2900/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3000/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3100/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3200/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3300/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3400/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3500/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3600/3750                   Loss D: 10.0000, loss G: -0.4899\n",
+            "Epoch [10/10] Batch 3700/3750                   Loss D: 10.0000, loss G: -0.4899\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "noise.data.normal_(0, 1)\n",
+        "fake = generator(noise)\n",
+        "\n",
+        "plt.figure(figsize=(6, 7))\n",
+        "for i in range(16):\n",
+        "    plt.subplot(4, 4, i + 1)\n",
+        "    plt.imshow(fake[i].detach().numpy().reshape(28, 28), cmap=plt.cm.Greys_r)\n",
+        "    plt.axis('off')"
+      ],
+      "metadata": {
+        "id": "R0_7sIllb150",
+        "outputId": "fcc9578e-31ea-4635-b043-69a299b008f8",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 401
+        }
+      },
+      "execution_count": 17,
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 432x504 with 16 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {
+            "needs_background": "light"
+          }
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "class Generator(nn.Module):\n",
+        "    def __init__(self):\n",
+        "        super(Generator, self).__init__()\n",
+        "        self.model = nn.Sequential( \n",
+        "            nn.Linear(config.noise_size+10, 200),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(200, 28*28),\n",
+        "            nn.Sigmoid())\n",
+        "        self.label_emb = nn.Embedding(10, 10)\n",
+        "\n",
+        "    def forward(self, x, labels):\n",
+        "        x = x.view(x.size(0), 50)\n",
+        "        gen_input = torch.cat((x, self.label_emb(labels)), 1)\n",
+        "        img = self.model(gen_input)\n",
+        "        \n",
+        "        return img.view(x.size(0), 28, 28)\n",
+        "    \n",
+        "class Discriminator(nn.Module):\n",
+        "    def __init__(self):\n",
+        "        super(Discriminator, self).__init__()\n",
+        "        self.model = nn.Sequential(\n",
+        "            nn.Linear(28*28+10, 200),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(200, 50),\n",
+        "            nn.ReLU(inplace=True),\n",
+        "            nn.Linear(50, 1), \n",
+        "            nn.Sigmoid())\n",
+        "        self.label_emb = nn.Embedding(10, 10)\n",
+        "        \n",
+        "    def forward(self, x, labels):\n",
+        "        #print(x.size())\n",
+        "        d_in = torch.cat((x.view(x.size(0), -1), self.label_emb(labels)), -1)\n",
+        "        validity = self.model(d_in)\n",
+        "        return validity"
+      ],
+      "metadata": {
+        "id": "DdRWVbAUb8Ro"
+      },
+      "execution_count": 18,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "generator = Generator()\n",
+        "discriminator = Discriminator()\n",
+        "\n",
+        "optim_G = optim.Adam(params=generator.parameters(), lr=0.0001)\n",
+        "optim_D = optim.Adam(params=discriminator.parameters(), lr=0.0001)\n",
+        "\n",
+        "criterion = nn.BCELoss()"
+      ],
+      "metadata": {
+        "id": "xDuN7IrIb-Ho"
+      },
+      "execution_count": 19,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size))\n",
+        "fixed_noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size).normal_(0, 1))\n",
+        "label = Variable(torch.FloatTensor(config.batch_size))"
+      ],
+      "metadata": {
+        "id": "MgHMDH25b_u_"
+      },
+      "execution_count": 20,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "ERRD_x = np.zeros(config.num_epochs)\n",
+        "ERRD_z = np.zeros(config.num_epochs)\n",
+        "ERRG = np.zeros(config.num_epochs)\n",
+        "N = len(dataloader)\n",
+        "\n",
+        "for epoch in range(config.num_epochs):\n",
+        "    for iteration, (images, cat) in enumerate(dataloader):\n",
+        "        ####### \n",
+        "        # Discriminator stage: maximize log(D(x)) + log(1 - D(G(z))) \n",
+        "        #######\n",
+        "        discriminator.zero_grad()\n",
+        "        \n",
+        "        # real\n",
+        "        label.data.fill_(real_label)\n",
+        "        input_data = images.view(images.shape[0], -1)\n",
+        "        output = discriminator(input_data, cat)\n",
+        "        errD_x = criterion(output.view(16), label)\n",
+        "        ERRD_x[epoch] += errD_x.item()\n",
+        "        errD_x.backward()\n",
+        "        \n",
+        "        # fake \n",
+        "        noise.data.normal_(0, 1)\n",
+        "        fake = generator(noise, cat)\n",
+        "        label.data.fill_(fake_label)\n",
+        "        output = discriminator(fake.detach(), cat).view(16)\n",
+        "        errD_z = criterion(output, label)\n",
+        "        ERRD_z[epoch] += errD_z.item()\n",
+        "        errD_z.backward()\n",
+        "        \n",
+        "        optim_D.step()\n",
+        "        ####### \n",
+        "        # Generator stage: maximize log(D(G(x))\n",
+        "        #######\n",
+        "        generator.zero_grad()\n",
+        "        label.data.fill_(real_label)\n",
+        "        output = discriminator(fake, cat).view(16)\n",
+        "        errG = criterion(output, label)\n",
+        "        ERRG[epoch] += errG.item()\n",
+        "        errG.backward()\n",
+        "        \n",
+        "        optim_G.step()\n",
+        "        \n",
+        "        if (iteration+1) % config.print_freq == 0:\n",
+        "            print('Epoch:{} Iter: {} errD_x: {:.2f} errD_z: {:.2f} errG: {:.2f}'.format(epoch+1,\n",
+        "                                                                                            iteration+1, \n",
+        "                                                                                            errD_x.item(),\n",
+        "                                                                                            errD_z.item(), \n",
+        "                                                                                            errG.item()))"
+      ],
+      "metadata": {
+        "id": "aFNWRhYlcCCG",
+        "outputId": "45d1d653-6484-46c7-db78-0e0fd507b9c2",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        }
+      },
+      "execution_count": 21,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Epoch:1 Iter: 100 errD_x: 0.16 errD_z: 0.13 errG: 2.16\n",
+            "Epoch:1 Iter: 200 errD_x: 0.24 errD_z: 0.27 errG: 2.17\n",
+            "Epoch:1 Iter: 300 errD_x: 0.30 errD_z: 0.15 errG: 2.16\n",
+            "Epoch:1 Iter: 400 errD_x: 0.06 errD_z: 0.13 errG: 2.24\n",
+            "Epoch:1 Iter: 500 errD_x: 0.16 errD_z: 0.10 errG: 2.50\n",
+            "Epoch:1 Iter: 600 errD_x: 0.08 errD_z: 0.10 errG: 2.41\n",
+            "Epoch:1 Iter: 700 errD_x: 0.10 errD_z: 0.15 errG: 2.24\n",
+            "Epoch:1 Iter: 800 errD_x: 0.11 errD_z: 0.13 errG: 2.42\n",
+            "Epoch:1 Iter: 900 errD_x: 0.18 errD_z: 0.14 errG: 2.24\n",
+            "Epoch:1 Iter: 1000 errD_x: 0.21 errD_z: 0.17 errG: 2.34\n",
+            "Epoch:1 Iter: 1100 errD_x: 0.23 errD_z: 0.17 errG: 2.54\n",
+            "Epoch:1 Iter: 1200 errD_x: 0.10 errD_z: 0.15 errG: 2.25\n",
+            "Epoch:1 Iter: 1300 errD_x: 0.10 errD_z: 0.11 errG: 3.08\n",
+            "Epoch:1 Iter: 1400 errD_x: 0.14 errD_z: 0.08 errG: 3.29\n",
+            "Epoch:1 Iter: 1500 errD_x: 0.10 errD_z: 0.13 errG: 3.01\n",
+            "Epoch:1 Iter: 1600 errD_x: 0.21 errD_z: 0.09 errG: 2.77\n",
+            "Epoch:1 Iter: 1700 errD_x: 0.23 errD_z: 0.14 errG: 3.32\n",
+            "Epoch:1 Iter: 1800 errD_x: 0.05 errD_z: 0.07 errG: 3.81\n",
+            "Epoch:1 Iter: 1900 errD_x: 0.05 errD_z: 0.11 errG: 3.04\n",
+            "Epoch:1 Iter: 2000 errD_x: 0.10 errD_z: 0.10 errG: 3.18\n",
+            "Epoch:1 Iter: 2100 errD_x: 0.33 errD_z: 0.14 errG: 3.13\n",
+            "Epoch:1 Iter: 2200 errD_x: 0.13 errD_z: 0.07 errG: 3.55\n",
+            "Epoch:1 Iter: 2300 errD_x: 0.36 errD_z: 0.01 errG: 4.63\n",
+            "Epoch:1 Iter: 2400 errD_x: 0.06 errD_z: 0.05 errG: 3.63\n",
+            "Epoch:1 Iter: 2500 errD_x: 0.13 errD_z: 0.06 errG: 3.41\n",
+            "Epoch:1 Iter: 2600 errD_x: 0.00 errD_z: 0.01 errG: 4.93\n",
+            "Epoch:1 Iter: 2700 errD_x: 0.01 errD_z: 0.04 errG: 4.41\n",
+            "Epoch:1 Iter: 2800 errD_x: 0.00 errD_z: 0.02 errG: 5.07\n",
+            "Epoch:1 Iter: 2900 errD_x: 0.01 errD_z: 0.03 errG: 4.10\n",
+            "Epoch:1 Iter: 3000 errD_x: 0.03 errD_z: 0.03 errG: 3.88\n",
+            "Epoch:1 Iter: 3100 errD_x: 0.00 errD_z: 0.01 errG: 5.10\n",
+            "Epoch:1 Iter: 3200 errD_x: 0.02 errD_z: 0.03 errG: 4.20\n",
+            "Epoch:1 Iter: 3300 errD_x: 0.05 errD_z: 0.06 errG: 3.94\n",
+            "Epoch:1 Iter: 3400 errD_x: 0.03 errD_z: 0.13 errG: 2.84\n",
+            "Epoch:1 Iter: 3500 errD_x: 0.29 errD_z: 0.03 errG: 4.48\n",
+            "Epoch:1 Iter: 3600 errD_x: 0.44 errD_z: 0.02 errG: 4.24\n",
+            "Epoch:1 Iter: 3700 errD_x: 0.06 errD_z: 0.09 errG: 3.60\n",
+            "Epoch:2 Iter: 100 errD_x: 0.16 errD_z: 0.10 errG: 2.73\n",
+            "Epoch:2 Iter: 200 errD_x: 0.08 errD_z: 0.05 errG: 3.42\n",
+            "Epoch:2 Iter: 300 errD_x: 0.06 errD_z: 0.03 errG: 4.10\n",
+            "Epoch:2 Iter: 400 errD_x: 0.05 errD_z: 0.04 errG: 3.94\n",
+            "Epoch:2 Iter: 500 errD_x: 0.05 errD_z: 0.06 errG: 3.73\n",
+            "Epoch:2 Iter: 600 errD_x: 0.08 errD_z: 0.02 errG: 4.76\n",
+            "Epoch:2 Iter: 700 errD_x: 0.02 errD_z: 0.12 errG: 3.19\n",
+            "Epoch:2 Iter: 800 errD_x: 0.02 errD_z: 0.04 errG: 5.00\n",
+            "Epoch:2 Iter: 900 errD_x: 0.09 errD_z: 0.09 errG: 3.20\n",
+            "Epoch:2 Iter: 1000 errD_x: 0.02 errD_z: 0.10 errG: 3.14\n",
+            "Epoch:2 Iter: 1100 errD_x: 0.11 errD_z: 0.06 errG: 3.50\n",
+            "Epoch:2 Iter: 1200 errD_x: 0.03 errD_z: 0.10 errG: 3.13\n",
+            "Epoch:2 Iter: 1300 errD_x: 0.01 errD_z: 0.10 errG: 2.95\n",
+            "Epoch:2 Iter: 1400 errD_x: 0.80 errD_z: 0.01 errG: 4.01\n",
+            "Epoch:2 Iter: 1500 errD_x: 0.02 errD_z: 0.07 errG: 3.38\n",
+            "Epoch:2 Iter: 1600 errD_x: 0.04 errD_z: 0.19 errG: 2.87\n",
+            "Epoch:2 Iter: 1700 errD_x: 0.15 errD_z: 0.02 errG: 4.85\n",
+            "Epoch:2 Iter: 1800 errD_x: 0.08 errD_z: 0.08 errG: 3.85\n",
+            "Epoch:2 Iter: 1900 errD_x: 0.07 errD_z: 0.08 errG: 3.65\n",
+            "Epoch:2 Iter: 2000 errD_x: 0.24 errD_z: 0.08 errG: 3.23\n",
+            "Epoch:2 Iter: 2100 errD_x: 0.14 errD_z: 0.06 errG: 3.12\n",
+            "Epoch:2 Iter: 2200 errD_x: 0.07 errD_z: 0.14 errG: 2.51\n",
+            "Epoch:2 Iter: 2300 errD_x: 0.09 errD_z: 0.12 errG: 3.07\n",
+            "Epoch:2 Iter: 2400 errD_x: 0.21 errD_z: 0.03 errG: 3.74\n",
+            "Epoch:2 Iter: 2500 errD_x: 0.26 errD_z: 0.03 errG: 3.57\n",
+            "Epoch:2 Iter: 2600 errD_x: 0.28 errD_z: 0.07 errG: 3.67\n",
+            "Epoch:2 Iter: 2700 errD_x: 0.29 errD_z: 0.06 errG: 3.23\n",
+            "Epoch:2 Iter: 2800 errD_x: 0.09 errD_z: 0.08 errG: 3.05\n",
+            "Epoch:2 Iter: 2900 errD_x: 0.15 errD_z: 0.07 errG: 3.49\n",
+            "Epoch:2 Iter: 3000 errD_x: 0.11 errD_z: 0.16 errG: 2.72\n",
+            "Epoch:2 Iter: 3100 errD_x: 0.25 errD_z: 0.13 errG: 2.62\n",
+            "Epoch:2 Iter: 3200 errD_x: 0.13 errD_z: 0.07 errG: 2.95\n",
+            "Epoch:2 Iter: 3300 errD_x: 0.00 errD_z: 0.31 errG: 2.84\n",
+            "Epoch:2 Iter: 3400 errD_x: 0.08 errD_z: 0.05 errG: 4.42\n",
+            "Epoch:2 Iter: 3500 errD_x: 0.36 errD_z: 0.09 errG: 3.82\n",
+            "Epoch:2 Iter: 3600 errD_x: 0.26 errD_z: 0.11 errG: 4.04\n",
+            "Epoch:2 Iter: 3700 errD_x: 0.21 errD_z: 0.08 errG: 3.53\n",
+            "Epoch:3 Iter: 100 errD_x: 0.31 errD_z: 0.02 errG: 4.95\n",
+            "Epoch:3 Iter: 200 errD_x: 0.09 errD_z: 0.05 errG: 3.32\n",
+            "Epoch:3 Iter: 300 errD_x: 0.38 errD_z: 0.02 errG: 4.88\n",
+            "Epoch:3 Iter: 400 errD_x: 0.19 errD_z: 0.23 errG: 1.98\n",
+            "Epoch:3 Iter: 500 errD_x: 0.02 errD_z: 0.13 errG: 3.26\n",
+            "Epoch:3 Iter: 600 errD_x: 0.03 errD_z: 0.30 errG: 2.62\n",
+            "Epoch:3 Iter: 700 errD_x: 0.53 errD_z: 0.08 errG: 3.20\n",
+            "Epoch:3 Iter: 800 errD_x: 0.01 errD_z: 0.13 errG: 3.82\n",
+            "Epoch:3 Iter: 900 errD_x: 0.10 errD_z: 0.03 errG: 3.92\n",
+            "Epoch:3 Iter: 1000 errD_x: 0.09 errD_z: 0.01 errG: 6.17\n",
+            "Epoch:3 Iter: 1100 errD_x: 0.02 errD_z: 0.12 errG: 3.69\n",
+            "Epoch:3 Iter: 1200 errD_x: 0.29 errD_z: 0.03 errG: 4.25\n",
+            "Epoch:3 Iter: 1300 errD_x: 0.19 errD_z: 0.10 errG: 3.62\n",
+            "Epoch:3 Iter: 1400 errD_x: 0.11 errD_z: 0.05 errG: 3.41\n",
+            "Epoch:3 Iter: 1500 errD_x: 0.03 errD_z: 0.11 errG: 3.43\n",
+            "Epoch:3 Iter: 1600 errD_x: 0.03 errD_z: 0.04 errG: 5.44\n",
+            "Epoch:3 Iter: 1700 errD_x: 0.07 errD_z: 0.02 errG: 4.61\n",
+            "Epoch:3 Iter: 1800 errD_x: 0.24 errD_z: 0.11 errG: 3.23\n",
+            "Epoch:3 Iter: 1900 errD_x: 0.01 errD_z: 0.04 errG: 3.98\n",
+            "Epoch:3 Iter: 2000 errD_x: 0.09 errD_z: 0.03 errG: 4.51\n",
+            "Epoch:3 Iter: 2100 errD_x: 0.34 errD_z: 0.09 errG: 3.93\n",
+            "Epoch:3 Iter: 2200 errD_x: 0.18 errD_z: 0.12 errG: 3.76\n",
+            "Epoch:3 Iter: 2300 errD_x: 0.01 errD_z: 0.08 errG: 3.45\n",
+            "Epoch:3 Iter: 2400 errD_x: 0.03 errD_z: 0.21 errG: 3.74\n",
+            "Epoch:3 Iter: 2500 errD_x: 0.17 errD_z: 0.09 errG: 3.83\n",
+            "Epoch:3 Iter: 2600 errD_x: 0.20 errD_z: 0.11 errG: 3.58\n",
+            "Epoch:3 Iter: 2700 errD_x: 0.00 errD_z: 0.23 errG: 3.92\n",
+            "Epoch:3 Iter: 2800 errD_x: 0.37 errD_z: 0.07 errG: 4.66\n",
+            "Epoch:3 Iter: 2900 errD_x: 0.43 errD_z: 0.03 errG: 6.16\n",
+            "Epoch:3 Iter: 3000 errD_x: 0.43 errD_z: 0.08 errG: 3.04\n",
+            "Epoch:3 Iter: 3100 errD_x: 0.06 errD_z: 0.06 errG: 4.03\n",
+            "Epoch:3 Iter: 3200 errD_x: 0.11 errD_z: 0.05 errG: 5.41\n",
+            "Epoch:3 Iter: 3300 errD_x: 0.03 errD_z: 0.03 errG: 5.59\n",
+            "Epoch:3 Iter: 3400 errD_x: 0.21 errD_z: 0.18 errG: 3.17\n",
+            "Epoch:3 Iter: 3500 errD_x: 0.40 errD_z: 0.02 errG: 5.90\n",
+            "Epoch:3 Iter: 3600 errD_x: 0.12 errD_z: 0.39 errG: 3.18\n",
+            "Epoch:3 Iter: 3700 errD_x: 0.31 errD_z: 0.13 errG: 5.65\n",
+            "Epoch:4 Iter: 100 errD_x: 0.01 errD_z: 0.09 errG: 4.55\n",
+            "Epoch:4 Iter: 200 errD_x: 0.06 errD_z: 0.14 errG: 2.88\n",
+            "Epoch:4 Iter: 300 errD_x: 0.29 errD_z: 0.17 errG: 2.31\n",
+            "Epoch:4 Iter: 400 errD_x: 0.07 errD_z: 0.02 errG: 6.55\n",
+            "Epoch:4 Iter: 500 errD_x: 0.02 errD_z: 0.09 errG: 4.23\n",
+            "Epoch:4 Iter: 600 errD_x: 0.37 errD_z: 0.07 errG: 4.18\n",
+            "Epoch:4 Iter: 700 errD_x: 0.49 errD_z: 0.10 errG: 4.67\n",
+            "Epoch:4 Iter: 800 errD_x: 0.07 errD_z: 0.09 errG: 4.58\n",
+            "Epoch:4 Iter: 900 errD_x: 0.05 errD_z: 0.08 errG: 4.25\n",
+            "Epoch:4 Iter: 1000 errD_x: 0.13 errD_z: 0.15 errG: 3.51\n",
+            "Epoch:4 Iter: 1100 errD_x: 0.06 errD_z: 0.17 errG: 3.06\n",
+            "Epoch:4 Iter: 1200 errD_x: 0.11 errD_z: 0.14 errG: 2.69\n",
+            "Epoch:4 Iter: 1300 errD_x: 0.13 errD_z: 0.15 errG: 3.84\n",
+            "Epoch:4 Iter: 1400 errD_x: 0.16 errD_z: 0.05 errG: 4.60\n",
+            "Epoch:4 Iter: 1500 errD_x: 0.42 errD_z: 0.10 errG: 4.93\n",
+            "Epoch:4 Iter: 1600 errD_x: 0.19 errD_z: 0.08 errG: 4.00\n",
+            "Epoch:4 Iter: 1700 errD_x: 0.08 errD_z: 0.04 errG: 5.56\n",
+            "Epoch:4 Iter: 1800 errD_x: 0.12 errD_z: 0.04 errG: 4.40\n",
+            "Epoch:4 Iter: 1900 errD_x: 0.06 errD_z: 0.02 errG: 4.88\n",
+            "Epoch:4 Iter: 2000 errD_x: 0.00 errD_z: 0.15 errG: 3.34\n",
+            "Epoch:4 Iter: 2100 errD_x: 0.25 errD_z: 0.13 errG: 4.03\n",
+            "Epoch:4 Iter: 2200 errD_x: 0.06 errD_z: 0.15 errG: 3.00\n",
+            "Epoch:4 Iter: 2300 errD_x: 0.31 errD_z: 0.05 errG: 3.57\n",
+            "Epoch:4 Iter: 2400 errD_x: 0.34 errD_z: 0.06 errG: 4.12\n",
+            "Epoch:4 Iter: 2500 errD_x: 0.11 errD_z: 0.19 errG: 3.01\n",
+            "Epoch:4 Iter: 2600 errD_x: 0.02 errD_z: 0.14 errG: 4.02\n",
+            "Epoch:4 Iter: 2700 errD_x: 0.38 errD_z: 0.09 errG: 4.42\n",
+            "Epoch:4 Iter: 2800 errD_x: 0.02 errD_z: 0.09 errG: 3.42\n",
+            "Epoch:4 Iter: 2900 errD_x: 0.11 errD_z: 0.10 errG: 3.75\n",
+            "Epoch:4 Iter: 3000 errD_x: 0.04 errD_z: 0.06 errG: 4.20\n",
+            "Epoch:4 Iter: 3100 errD_x: 0.20 errD_z: 0.03 errG: 4.83\n",
+            "Epoch:4 Iter: 3200 errD_x: 0.27 errD_z: 0.05 errG: 5.26\n",
+            "Epoch:4 Iter: 3300 errD_x: 0.03 errD_z: 0.03 errG: 5.01\n",
+            "Epoch:4 Iter: 3400 errD_x: 0.14 errD_z: 0.13 errG: 4.06\n",
+            "Epoch:4 Iter: 3500 errD_x: 0.01 errD_z: 0.05 errG: 4.44\n",
+            "Epoch:4 Iter: 3600 errD_x: 0.08 errD_z: 0.12 errG: 3.53\n",
+            "Epoch:4 Iter: 3700 errD_x: 0.30 errD_z: 0.08 errG: 3.37\n",
+            "Epoch:5 Iter: 100 errD_x: 0.09 errD_z: 0.11 errG: 4.23\n",
+            "Epoch:5 Iter: 200 errD_x: 0.44 errD_z: 0.12 errG: 2.91\n",
+            "Epoch:5 Iter: 300 errD_x: 0.08 errD_z: 0.14 errG: 3.42\n",
+            "Epoch:5 Iter: 400 errD_x: 0.30 errD_z: 0.09 errG: 3.99\n",
+            "Epoch:5 Iter: 500 errD_x: 0.04 errD_z: 0.06 errG: 4.38\n",
+            "Epoch:5 Iter: 600 errD_x: 0.12 errD_z: 0.09 errG: 4.47\n",
+            "Epoch:5 Iter: 700 errD_x: 0.11 errD_z: 0.10 errG: 3.98\n",
+            "Epoch:5 Iter: 800 errD_x: 0.02 errD_z: 0.02 errG: 4.61\n",
+            "Epoch:5 Iter: 900 errD_x: 0.25 errD_z: 0.16 errG: 4.23\n",
+            "Epoch:5 Iter: 1000 errD_x: 0.19 errD_z: 0.04 errG: 4.87\n",
+            "Epoch:5 Iter: 1100 errD_x: 0.60 errD_z: 0.09 errG: 3.72\n",
+            "Epoch:5 Iter: 1200 errD_x: 0.14 errD_z: 0.18 errG: 4.01\n",
+            "Epoch:5 Iter: 1300 errD_x: 0.32 errD_z: 0.03 errG: 4.82\n",
+            "Epoch:5 Iter: 1400 errD_x: 0.10 errD_z: 0.13 errG: 4.17\n",
+            "Epoch:5 Iter: 1500 errD_x: 0.01 errD_z: 0.35 errG: 3.44\n",
+            "Epoch:5 Iter: 1600 errD_x: 0.13 errD_z: 0.04 errG: 5.92\n",
+            "Epoch:5 Iter: 1700 errD_x: 0.07 errD_z: 0.25 errG: 4.99\n",
+            "Epoch:5 Iter: 1800 errD_x: 0.11 errD_z: 0.20 errG: 3.61\n",
+            "Epoch:5 Iter: 1900 errD_x: 0.39 errD_z: 0.04 errG: 4.30\n",
+            "Epoch:5 Iter: 2000 errD_x: 0.40 errD_z: 0.11 errG: 4.59\n",
+            "Epoch:5 Iter: 2100 errD_x: 0.20 errD_z: 0.11 errG: 3.25\n",
+            "Epoch:5 Iter: 2200 errD_x: 0.22 errD_z: 0.03 errG: 4.73\n",
+            "Epoch:5 Iter: 2300 errD_x: 0.09 errD_z: 0.20 errG: 2.99\n",
+            "Epoch:5 Iter: 2400 errD_x: 0.36 errD_z: 0.09 errG: 3.72\n",
+            "Epoch:5 Iter: 2500 errD_x: 0.34 errD_z: 0.12 errG: 3.28\n",
+            "Epoch:5 Iter: 2600 errD_x: 0.04 errD_z: 0.23 errG: 3.20\n",
+            "Epoch:5 Iter: 2700 errD_x: 0.16 errD_z: 0.01 errG: 6.29\n",
+            "Epoch:5 Iter: 2800 errD_x: 0.28 errD_z: 0.27 errG: 3.01\n",
+            "Epoch:5 Iter: 2900 errD_x: 0.06 errD_z: 0.09 errG: 4.00\n",
+            "Epoch:5 Iter: 3000 errD_x: 0.01 errD_z: 0.08 errG: 5.49\n",
+            "Epoch:5 Iter: 3100 errD_x: 0.26 errD_z: 0.03 errG: 4.47\n",
+            "Epoch:5 Iter: 3200 errD_x: 0.08 errD_z: 0.19 errG: 4.39\n",
+            "Epoch:5 Iter: 3300 errD_x: 0.60 errD_z: 0.23 errG: 3.08\n",
+            "Epoch:5 Iter: 3400 errD_x: 0.68 errD_z: 0.33 errG: 2.69\n",
+            "Epoch:5 Iter: 3500 errD_x: 0.40 errD_z: 0.13 errG: 3.24\n",
+            "Epoch:5 Iter: 3600 errD_x: 0.94 errD_z: 0.53 errG: 3.15\n",
+            "Epoch:5 Iter: 3700 errD_x: 0.08 errD_z: 0.32 errG: 2.38\n",
+            "Epoch:6 Iter: 100 errD_x: 0.30 errD_z: 0.07 errG: 3.37\n",
+            "Epoch:6 Iter: 200 errD_x: 0.21 errD_z: 0.03 errG: 4.47\n",
+            "Epoch:6 Iter: 300 errD_x: 0.56 errD_z: 0.02 errG: 5.40\n",
+            "Epoch:6 Iter: 400 errD_x: 0.29 errD_z: 0.08 errG: 4.51\n",
+            "Epoch:6 Iter: 500 errD_x: 0.17 errD_z: 0.14 errG: 3.26\n",
+            "Epoch:6 Iter: 600 errD_x: 0.29 errD_z: 0.06 errG: 4.42\n",
+            "Epoch:6 Iter: 700 errD_x: 0.34 errD_z: 0.20 errG: 3.07\n",
+            "Epoch:6 Iter: 800 errD_x: 0.47 errD_z: 0.10 errG: 3.63\n",
+            "Epoch:6 Iter: 900 errD_x: 0.32 errD_z: 0.09 errG: 4.01\n",
+            "Epoch:6 Iter: 1000 errD_x: 0.03 errD_z: 0.23 errG: 3.07\n",
+            "Epoch:6 Iter: 1100 errD_x: 0.11 errD_z: 0.11 errG: 3.42\n",
+            "Epoch:6 Iter: 1200 errD_x: 0.57 errD_z: 0.17 errG: 2.96\n",
+            "Epoch:6 Iter: 1300 errD_x: 0.27 errD_z: 0.11 errG: 3.75\n",
+            "Epoch:6 Iter: 1400 errD_x: 0.01 errD_z: 0.08 errG: 5.18\n",
+            "Epoch:6 Iter: 1500 errD_x: 0.05 errD_z: 0.14 errG: 3.43\n",
+            "Epoch:6 Iter: 1600 errD_x: 0.34 errD_z: 0.02 errG: 4.42\n",
+            "Epoch:6 Iter: 1700 errD_x: 0.08 errD_z: 0.17 errG: 2.35\n",
+            "Epoch:6 Iter: 1800 errD_x: 0.04 errD_z: 0.17 errG: 2.75\n",
+            "Epoch:6 Iter: 1900 errD_x: 0.04 errD_z: 0.35 errG: 2.96\n",
+            "Epoch:6 Iter: 2000 errD_x: 0.05 errD_z: 0.16 errG: 2.73\n",
+            "Epoch:6 Iter: 2100 errD_x: 0.44 errD_z: 0.06 errG: 3.73\n",
+            "Epoch:6 Iter: 2200 errD_x: 0.65 errD_z: 0.18 errG: 2.88\n",
+            "Epoch:6 Iter: 2300 errD_x: 0.52 errD_z: 0.32 errG: 2.98\n",
+            "Epoch:6 Iter: 2400 errD_x: 0.05 errD_z: 0.36 errG: 3.56\n",
+            "Epoch:6 Iter: 2500 errD_x: 0.43 errD_z: 0.17 errG: 3.36\n",
+            "Epoch:6 Iter: 2600 errD_x: 0.87 errD_z: 0.07 errG: 3.76\n",
+            "Epoch:6 Iter: 2700 errD_x: 0.69 errD_z: 0.11 errG: 3.19\n",
+            "Epoch:6 Iter: 2800 errD_x: 0.25 errD_z: 0.39 errG: 2.59\n",
+            "Epoch:6 Iter: 2900 errD_x: 0.29 errD_z: 0.17 errG: 3.68\n",
+            "Epoch:6 Iter: 3000 errD_x: 0.48 errD_z: 0.41 errG: 2.15\n",
+            "Epoch:6 Iter: 3100 errD_x: 0.28 errD_z: 0.32 errG: 2.65\n",
+            "Epoch:6 Iter: 3200 errD_x: 0.12 errD_z: 0.08 errG: 3.59\n",
+            "Epoch:6 Iter: 3300 errD_x: 0.06 errD_z: 0.10 errG: 3.54\n",
+            "Epoch:6 Iter: 3400 errD_x: 0.01 errD_z: 0.43 errG: 3.03\n",
+            "Epoch:6 Iter: 3500 errD_x: 0.24 errD_z: 0.32 errG: 2.28\n",
+            "Epoch:6 Iter: 3600 errD_x: 0.01 errD_z: 0.17 errG: 3.77\n",
+            "Epoch:6 Iter: 3700 errD_x: 0.02 errD_z: 0.33 errG: 2.98\n",
+            "Epoch:7 Iter: 100 errD_x: 0.11 errD_z: 0.55 errG: 2.49\n",
+            "Epoch:7 Iter: 200 errD_x: 0.09 errD_z: 0.22 errG: 2.97\n",
+            "Epoch:7 Iter: 300 errD_x: 0.10 errD_z: 0.09 errG: 3.83\n",
+            "Epoch:7 Iter: 400 errD_x: 0.30 errD_z: 0.10 errG: 4.01\n",
+            "Epoch:7 Iter: 500 errD_x: 0.29 errD_z: 0.23 errG: 2.88\n",
+            "Epoch:7 Iter: 600 errD_x: 0.08 errD_z: 0.14 errG: 3.14\n",
+            "Epoch:7 Iter: 700 errD_x: 0.12 errD_z: 0.34 errG: 2.40\n",
+            "Epoch:7 Iter: 800 errD_x: 0.22 errD_z: 0.17 errG: 2.87\n",
+            "Epoch:7 Iter: 900 errD_x: 0.05 errD_z: 0.17 errG: 3.75\n",
+            "Epoch:7 Iter: 1000 errD_x: 0.12 errD_z: 0.30 errG: 2.88\n",
+            "Epoch:7 Iter: 1100 errD_x: 0.17 errD_z: 0.09 errG: 3.61\n",
+            "Epoch:7 Iter: 1200 errD_x: 0.08 errD_z: 0.49 errG: 1.94\n",
+            "Epoch:7 Iter: 1300 errD_x: 0.81 errD_z: 0.17 errG: 3.21\n",
+            "Epoch:7 Iter: 1400 errD_x: 0.23 errD_z: 0.31 errG: 2.52\n",
+            "Epoch:7 Iter: 1500 errD_x: 0.14 errD_z: 0.20 errG: 3.66\n",
+            "Epoch:7 Iter: 1600 errD_x: 0.50 errD_z: 0.26 errG: 2.16\n",
+            "Epoch:7 Iter: 1700 errD_x: 0.12 errD_z: 0.08 errG: 3.74\n",
+            "Epoch:7 Iter: 1800 errD_x: 0.23 errD_z: 0.13 errG: 3.05\n",
+            "Epoch:7 Iter: 1900 errD_x: 0.17 errD_z: 0.19 errG: 2.65\n",
+            "Epoch:7 Iter: 2000 errD_x: 0.45 errD_z: 0.24 errG: 2.73\n",
+            "Epoch:7 Iter: 2100 errD_x: 0.07 errD_z: 0.19 errG: 4.38\n",
+            "Epoch:7 Iter: 2200 errD_x: 0.14 errD_z: 0.07 errG: 4.95\n",
+            "Epoch:7 Iter: 2300 errD_x: 0.16 errD_z: 0.14 errG: 2.61\n",
+            "Epoch:7 Iter: 2400 errD_x: 0.69 errD_z: 0.12 errG: 4.39\n",
+            "Epoch:7 Iter: 2500 errD_x: 0.55 errD_z: 0.03 errG: 5.13\n",
+            "Epoch:7 Iter: 2600 errD_x: 0.66 errD_z: 0.20 errG: 2.59\n",
+            "Epoch:7 Iter: 2700 errD_x: 0.62 errD_z: 0.17 errG: 3.08\n",
+            "Epoch:7 Iter: 2800 errD_x: 0.49 errD_z: 0.36 errG: 2.56\n",
+            "Epoch:7 Iter: 2900 errD_x: 0.42 errD_z: 0.07 errG: 3.65\n",
+            "Epoch:7 Iter: 3000 errD_x: 0.42 errD_z: 0.11 errG: 3.37\n",
+            "Epoch:7 Iter: 3100 errD_x: 0.09 errD_z: 0.01 errG: 5.10\n",
+            "Epoch:7 Iter: 3200 errD_x: 0.23 errD_z: 0.09 errG: 4.31\n",
+            "Epoch:7 Iter: 3300 errD_x: 0.40 errD_z: 0.14 errG: 2.80\n",
+            "Epoch:7 Iter: 3400 errD_x: 0.01 errD_z: 0.29 errG: 3.83\n",
+            "Epoch:7 Iter: 3500 errD_x: 0.69 errD_z: 0.10 errG: 2.58\n",
+            "Epoch:7 Iter: 3600 errD_x: 0.15 errD_z: 0.24 errG: 3.52\n",
+            "Epoch:7 Iter: 3700 errD_x: 0.11 errD_z: 0.16 errG: 2.97\n",
+            "Epoch:8 Iter: 100 errD_x: 0.05 errD_z: 0.21 errG: 3.88\n",
+            "Epoch:8 Iter: 200 errD_x: 0.04 errD_z: 0.45 errG: 3.28\n",
+            "Epoch:8 Iter: 300 errD_x: 0.05 errD_z: 0.09 errG: 3.70\n",
+            "Epoch:8 Iter: 400 errD_x: 0.10 errD_z: 0.16 errG: 4.28\n",
+            "Epoch:8 Iter: 500 errD_x: 0.04 errD_z: 0.39 errG: 2.10\n",
+            "Epoch:8 Iter: 600 errD_x: 0.50 errD_z: 0.22 errG: 3.36\n",
+            "Epoch:8 Iter: 700 errD_x: 0.61 errD_z: 0.07 errG: 3.25\n",
+            "Epoch:8 Iter: 800 errD_x: 0.12 errD_z: 0.08 errG: 3.75\n",
+            "Epoch:8 Iter: 900 errD_x: 0.16 errD_z: 0.15 errG: 2.88\n",
+            "Epoch:8 Iter: 1000 errD_x: 0.42 errD_z: 0.12 errG: 3.86\n",
+            "Epoch:8 Iter: 1100 errD_x: 0.56 errD_z: 0.02 errG: 5.33\n",
+            "Epoch:8 Iter: 1200 errD_x: 0.12 errD_z: 0.08 errG: 3.52\n",
+            "Epoch:8 Iter: 1300 errD_x: 0.66 errD_z: 0.04 errG: 3.51\n",
+            "Epoch:8 Iter: 1400 errD_x: 0.19 errD_z: 0.20 errG: 2.79\n",
+            "Epoch:8 Iter: 1500 errD_x: 0.72 errD_z: 0.14 errG: 4.13\n",
+            "Epoch:8 Iter: 1600 errD_x: 0.15 errD_z: 0.09 errG: 3.76\n",
+            "Epoch:8 Iter: 1700 errD_x: 0.22 errD_z: 0.13 errG: 4.03\n",
+            "Epoch:8 Iter: 1800 errD_x: 0.55 errD_z: 0.15 errG: 2.47\n",
+            "Epoch:8 Iter: 1900 errD_x: 0.24 errD_z: 0.42 errG: 3.09\n",
+            "Epoch:8 Iter: 2000 errD_x: 0.25 errD_z: 0.11 errG: 3.51\n",
+            "Epoch:8 Iter: 2100 errD_x: 0.51 errD_z: 0.66 errG: 1.51\n",
+            "Epoch:8 Iter: 2200 errD_x: 0.05 errD_z: 0.14 errG: 2.78\n",
+            "Epoch:8 Iter: 2300 errD_x: 0.10 errD_z: 0.23 errG: 2.44\n",
+            "Epoch:8 Iter: 2400 errD_x: 0.43 errD_z: 0.12 errG: 4.11\n",
+            "Epoch:8 Iter: 2500 errD_x: 0.43 errD_z: 0.05 errG: 4.38\n",
+            "Epoch:8 Iter: 2600 errD_x: 0.24 errD_z: 0.10 errG: 3.13\n",
+            "Epoch:8 Iter: 2700 errD_x: 0.05 errD_z: 0.10 errG: 3.80\n",
+            "Epoch:8 Iter: 2800 errD_x: 0.34 errD_z: 0.12 errG: 2.68\n",
+            "Epoch:8 Iter: 2900 errD_x: 0.39 errD_z: 0.27 errG: 2.20\n",
+            "Epoch:8 Iter: 3000 errD_x: 0.30 errD_z: 0.21 errG: 2.50\n",
+            "Epoch:8 Iter: 3100 errD_x: 0.38 errD_z: 0.17 errG: 3.60\n",
+            "Epoch:8 Iter: 3200 errD_x: 0.46 errD_z: 0.12 errG: 2.65\n",
+            "Epoch:8 Iter: 3300 errD_x: 0.19 errD_z: 0.25 errG: 2.53\n",
+            "Epoch:8 Iter: 3400 errD_x: 0.17 errD_z: 0.14 errG: 2.98\n",
+            "Epoch:8 Iter: 3500 errD_x: 0.12 errD_z: 0.13 errG: 3.36\n",
+            "Epoch:8 Iter: 3600 errD_x: 0.09 errD_z: 0.09 errG: 3.11\n",
+            "Epoch:8 Iter: 3700 errD_x: 0.05 errD_z: 0.13 errG: 3.54\n",
+            "Epoch:9 Iter: 100 errD_x: 0.19 errD_z: 0.35 errG: 1.81\n",
+            "Epoch:9 Iter: 200 errD_x: 0.15 errD_z: 0.05 errG: 4.40\n",
+            "Epoch:9 Iter: 300 errD_x: 0.37 errD_z: 0.15 errG: 3.27\n",
+            "Epoch:9 Iter: 400 errD_x: 0.07 errD_z: 0.10 errG: 3.49\n",
+            "Epoch:9 Iter: 500 errD_x: 0.30 errD_z: 0.04 errG: 4.66\n",
+            "Epoch:9 Iter: 600 errD_x: 0.46 errD_z: 0.39 errG: 2.10\n",
+            "Epoch:9 Iter: 700 errD_x: 0.36 errD_z: 0.43 errG: 2.36\n",
+            "Epoch:9 Iter: 800 errD_x: 0.02 errD_z: 0.21 errG: 2.67\n",
+            "Epoch:9 Iter: 900 errD_x: 0.27 errD_z: 0.06 errG: 3.67\n",
+            "Epoch:9 Iter: 1000 errD_x: 0.69 errD_z: 0.12 errG: 4.01\n",
+            "Epoch:9 Iter: 1100 errD_x: 0.04 errD_z: 0.49 errG: 2.68\n",
+            "Epoch:9 Iter: 1200 errD_x: 0.08 errD_z: 0.17 errG: 3.11\n",
+            "Epoch:9 Iter: 1300 errD_x: 0.13 errD_z: 0.27 errG: 2.67\n",
+            "Epoch:9 Iter: 1400 errD_x: 0.08 errD_z: 0.36 errG: 1.98\n",
+            "Epoch:9 Iter: 1500 errD_x: 0.28 errD_z: 0.25 errG: 3.46\n",
+            "Epoch:9 Iter: 1600 errD_x: 0.75 errD_z: 0.09 errG: 2.58\n",
+            "Epoch:9 Iter: 1700 errD_x: 0.26 errD_z: 0.10 errG: 3.31\n",
+            "Epoch:9 Iter: 1800 errD_x: 0.28 errD_z: 0.17 errG: 3.34\n",
+            "Epoch:9 Iter: 1900 errD_x: 0.21 errD_z: 0.07 errG: 3.31\n",
+            "Epoch:9 Iter: 2000 errD_x: 0.35 errD_z: 0.24 errG: 2.73\n",
+            "Epoch:9 Iter: 2100 errD_x: 0.32 errD_z: 0.26 errG: 3.89\n",
+            "Epoch:9 Iter: 2200 errD_x: 0.08 errD_z: 0.33 errG: 2.79\n",
+            "Epoch:9 Iter: 2300 errD_x: 0.50 errD_z: 0.24 errG: 1.94\n",
+            "Epoch:9 Iter: 2400 errD_x: 0.12 errD_z: 0.53 errG: 2.21\n",
+            "Epoch:9 Iter: 2500 errD_x: 0.17 errD_z: 0.10 errG: 3.28\n",
+            "Epoch:9 Iter: 2600 errD_x: 0.19 errD_z: 0.30 errG: 2.54\n",
+            "Epoch:9 Iter: 2700 errD_x: 1.01 errD_z: 0.07 errG: 3.95\n",
+            "Epoch:9 Iter: 2800 errD_x: 0.11 errD_z: 0.31 errG: 2.29\n",
+            "Epoch:9 Iter: 2900 errD_x: 0.85 errD_z: 0.36 errG: 2.27\n",
+            "Epoch:9 Iter: 3000 errD_x: 0.30 errD_z: 0.14 errG: 2.54\n",
+            "Epoch:9 Iter: 3100 errD_x: 0.35 errD_z: 0.24 errG: 2.68\n",
+            "Epoch:9 Iter: 3200 errD_x: 0.35 errD_z: 0.48 errG: 1.83\n",
+            "Epoch:9 Iter: 3300 errD_x: 0.12 errD_z: 0.34 errG: 2.85\n",
+            "Epoch:9 Iter: 3400 errD_x: 0.10 errD_z: 0.18 errG: 2.96\n",
+            "Epoch:9 Iter: 3500 errD_x: 0.06 errD_z: 0.20 errG: 2.67\n",
+            "Epoch:9 Iter: 3600 errD_x: 0.36 errD_z: 0.81 errG: 1.83\n",
+            "Epoch:9 Iter: 3700 errD_x: 0.57 errD_z: 0.24 errG: 2.58\n",
+            "Epoch:10 Iter: 100 errD_x: 0.26 errD_z: 0.11 errG: 2.88\n",
+            "Epoch:10 Iter: 200 errD_x: 0.36 errD_z: 0.23 errG: 2.52\n",
+            "Epoch:10 Iter: 300 errD_x: 0.49 errD_z: 0.04 errG: 4.18\n",
+            "Epoch:10 Iter: 400 errD_x: 0.55 errD_z: 0.11 errG: 3.44\n",
+            "Epoch:10 Iter: 500 errD_x: 0.61 errD_z: 0.13 errG: 2.57\n",
+            "Epoch:10 Iter: 600 errD_x: 0.21 errD_z: 0.50 errG: 2.63\n",
+            "Epoch:10 Iter: 700 errD_x: 0.32 errD_z: 0.25 errG: 3.34\n",
+            "Epoch:10 Iter: 800 errD_x: 0.25 errD_z: 0.23 errG: 3.42\n",
+            "Epoch:10 Iter: 900 errD_x: 0.36 errD_z: 0.07 errG: 2.95\n",
+            "Epoch:10 Iter: 1000 errD_x: 0.21 errD_z: 0.06 errG: 4.03\n",
+            "Epoch:10 Iter: 1100 errD_x: 0.24 errD_z: 0.38 errG: 3.23\n",
+            "Epoch:10 Iter: 1200 errD_x: 0.07 errD_z: 0.21 errG: 3.65\n",
+            "Epoch:10 Iter: 1300 errD_x: 0.34 errD_z: 0.22 errG: 2.63\n",
+            "Epoch:10 Iter: 1400 errD_x: 0.59 errD_z: 0.23 errG: 2.73\n",
+            "Epoch:10 Iter: 1500 errD_x: 0.26 errD_z: 0.40 errG: 1.74\n",
+            "Epoch:10 Iter: 1600 errD_x: 0.39 errD_z: 0.23 errG: 2.08\n",
+            "Epoch:10 Iter: 1700 errD_x: 0.21 errD_z: 0.23 errG: 2.66\n",
+            "Epoch:10 Iter: 1800 errD_x: 0.38 errD_z: 0.06 errG: 4.06\n",
+            "Epoch:10 Iter: 1900 errD_x: 0.20 errD_z: 0.14 errG: 2.94\n",
+            "Epoch:10 Iter: 2000 errD_x: 0.10 errD_z: 0.39 errG: 2.28\n",
+            "Epoch:10 Iter: 2100 errD_x: 0.58 errD_z: 0.07 errG: 3.03\n",
+            "Epoch:10 Iter: 2200 errD_x: 0.31 errD_z: 0.13 errG: 3.29\n",
+            "Epoch:10 Iter: 2300 errD_x: 0.39 errD_z: 0.19 errG: 2.37\n",
+            "Epoch:10 Iter: 2400 errD_x: 0.88 errD_z: 0.13 errG: 2.63\n",
+            "Epoch:10 Iter: 2500 errD_x: 1.10 errD_z: 0.15 errG: 3.05\n",
+            "Epoch:10 Iter: 2600 errD_x: 0.04 errD_z: 0.09 errG: 3.77\n",
+            "Epoch:10 Iter: 2700 errD_x: 0.14 errD_z: 0.13 errG: 3.54\n",
+            "Epoch:10 Iter: 2800 errD_x: 0.36 errD_z: 0.41 errG: 2.22\n",
+            "Epoch:10 Iter: 2900 errD_x: 0.18 errD_z: 0.29 errG: 2.27\n",
+            "Epoch:10 Iter: 3000 errD_x: 0.44 errD_z: 0.19 errG: 2.31\n",
+            "Epoch:10 Iter: 3100 errD_x: 0.25 errD_z: 0.16 errG: 3.19\n",
+            "Epoch:10 Iter: 3200 errD_x: 0.57 errD_z: 0.27 errG: 3.00\n",
+            "Epoch:10 Iter: 3300 errD_x: 0.29 errD_z: 0.15 errG: 2.92\n",
+            "Epoch:10 Iter: 3400 errD_x: 0.07 errD_z: 0.20 errG: 2.62\n",
+            "Epoch:10 Iter: 3500 errD_x: 0.06 errD_z: 0.33 errG: 1.95\n",
+            "Epoch:10 Iter: 3600 errD_x: 0.21 errD_z: 0.19 errG: 2.85\n",
+            "Epoch:10 Iter: 3700 errD_x: 0.37 errD_z: 0.19 errG: 2.22\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "noise.data.normal_(0, 1)\n",
+        "fake = generator(noise, cat)\n",
+        "\n",
+        "plt.figure(figsize=(6, 7))\n",
+        "for i in range(16):\n",
+        "    plt.subplot(4, 4, i + 1)\n",
+        "    plt.imshow(fake[i].detach().numpy().reshape(28, 28), cmap=plt.cm.Greys_r)\n",
+        "    plt.axis('off')"
+      ],
+      "metadata": {
+        "id": "JyK5AzildkY1",
+        "outputId": "6eb2e175-e56a-41d4-ab26-a714dd7c2c72",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 401
+        }
+      },
+      "execution_count": 22,
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 432x504 with 16 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {
+            "needs_background": "light"
+          }
+        }
+      ]
+    }
+  ],
+  "metadata": {
+    "anaconda-cloud": {},
+    "kernelspec": {
+      "display_name": "Python 3 (ipykernel)",
+      "language": "python",
+      "name": "python3"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3.8.11"
+    },
+    "colab": {
+      "provenance": []
+    },
+    "widgets": {
+      "application/vnd.jupyter.widget-state+json": {
+        "3089611b5cda418faf550134695bd1c2": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HBoxModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_02e561b7133b4c5e920a3a2c0d300d2e",
+              "IPY_MODEL_d25fd1b9a05f404eb37505b593c83bee",
+              "IPY_MODEL_70d8400cd7474cf3984d1f31eb3f6fd7"
+            ],
+            "layout": "IPY_MODEL_ab3ec40e6808496dbb5e46d77773eff7"
+          }
+        },
+        "02e561b7133b4c5e920a3a2c0d300d2e": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_993857a0d4574ff8bef19a791c0b9e3c",
+            "placeholder": "​",
+            "style": "IPY_MODEL_00258c5fbf8348a79f9d859a34a5df43",
+            "value": "100%"
+          }
+        },
+        "d25fd1b9a05f404eb37505b593c83bee": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "FloatProgressModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_37d0caea6f5146c98eb549c98b113d9a",
+            "max": 26421880,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_26a20977835d4f8ea9dbc0f06427bdbb",
+            "value": 26421880
+          }
+        },
+        "70d8400cd7474cf3984d1f31eb3f6fd7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_8c987b2fd7c249e7b74623c91a3f15ce",
+            "placeholder": "​",
+            "style": "IPY_MODEL_fb3036cf1cd0437b86ac1632219f7e04",
+            "value": " 26421880/26421880 [00:01&lt;00:00, 29775105.97it/s]"
+          }
+        },
+        "ab3ec40e6808496dbb5e46d77773eff7": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "993857a0d4574ff8bef19a791c0b9e3c": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "00258c5fbf8348a79f9d859a34a5df43": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "37d0caea6f5146c98eb549c98b113d9a": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "26a20977835d4f8ea9dbc0f06427bdbb": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "ProgressStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "8c987b2fd7c249e7b74623c91a3f15ce": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "fb3036cf1cd0437b86ac1632219f7e04": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "65be345916ef419893271ff1f0a92b51": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HBoxModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_37c36bd250e04124a09270e46141f41f",
+              "IPY_MODEL_2937ae2f576b40ada1f51eac29d28422",
+              "IPY_MODEL_f68051dc0ef54f0f978d67f303a22e9d"
+            ],
+            "layout": "IPY_MODEL_dd0cbf60b58149b89f0cd0e59c130b0e"
+          }
+        },
+        "37c36bd250e04124a09270e46141f41f": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_7a2f922f6aae441380f1586048914365",
+            "placeholder": "​",
+            "style": "IPY_MODEL_933951cefdc84b7796526bdc6869fdf3",
+            "value": "100%"
+          }
+        },
+        "2937ae2f576b40ada1f51eac29d28422": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "FloatProgressModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_3ff4afbc672c40d0b096f119a07d0024",
+            "max": 29515,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_b02746fe4d774ca39040f7685f7827cf",
+            "value": 29515
+          }
+        },
+        "f68051dc0ef54f0f978d67f303a22e9d": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_701dc00c6a24482192ecf3f082babc2c",
+            "placeholder": "​",
+            "style": "IPY_MODEL_c6547b6dabf148508e341c2bdf369d83",
+            "value": " 29515/29515 [00:00&lt;00:00, 331643.48it/s]"
+          }
+        },
+        "dd0cbf60b58149b89f0cd0e59c130b0e": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "7a2f922f6aae441380f1586048914365": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "933951cefdc84b7796526bdc6869fdf3": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "3ff4afbc672c40d0b096f119a07d0024": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "b02746fe4d774ca39040f7685f7827cf": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "ProgressStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "701dc00c6a24482192ecf3f082babc2c": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "c6547b6dabf148508e341c2bdf369d83": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "92cefc1774de486091dc3b60021783c1": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HBoxModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_8fa40adeb3cb428298430949d13ad090",
+              "IPY_MODEL_c3ab1c12ac574a15920daf69c738c302",
+              "IPY_MODEL_5a8a21e352be4118a88e75141ac843cc"
+            ],
+            "layout": "IPY_MODEL_37b2250444864bf880421d4d2fdc57ae"
+          }
+        },
+        "8fa40adeb3cb428298430949d13ad090": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_2d9f91fe44054b9ca60f153878eddcba",
+            "placeholder": "​",
+            "style": "IPY_MODEL_f271a7fe389546edbeddc0cee11444f4",
+            "value": "100%"
+          }
+        },
+        "c3ab1c12ac574a15920daf69c738c302": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "FloatProgressModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_ba1d248606e54691a73a062a9f5646cb",
+            "max": 4422102,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_340311831d5b492e99c0a5e015975443",
+            "value": 4422102
+          }
+        },
+        "5a8a21e352be4118a88e75141ac843cc": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_2f0bf104af174a21aecc2a320cece5c5",
+            "placeholder": "​",
+            "style": "IPY_MODEL_3bfc8f6bf8a6430693e0703b22a57272",
+            "value": " 4422102/4422102 [00:00&lt;00:00, 7045540.43it/s]"
+          }
+        },
+        "37b2250444864bf880421d4d2fdc57ae": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "2d9f91fe44054b9ca60f153878eddcba": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "f271a7fe389546edbeddc0cee11444f4": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "ba1d248606e54691a73a062a9f5646cb": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "340311831d5b492e99c0a5e015975443": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "ProgressStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "2f0bf104af174a21aecc2a320cece5c5": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "3bfc8f6bf8a6430693e0703b22a57272": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "e251c830789e469888cc2173200a66ff": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HBoxModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HBoxModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HBoxView",
+            "box_style": "",
+            "children": [
+              "IPY_MODEL_a6c11c59a074439cb9417b0b3047ff37",
+              "IPY_MODEL_4e0ce38999fd431c971f095a11493445",
+              "IPY_MODEL_5daff87ee00f4024a4d0c36ec365656d"
+            ],
+            "layout": "IPY_MODEL_89306a31ada645e887f280483191d061"
+          }
+        },
+        "a6c11c59a074439cb9417b0b3047ff37": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_83b00e40330e4a9bab89fce2d40aab7f",
+            "placeholder": "​",
+            "style": "IPY_MODEL_77b51cbd87e04728a058d46ee733c819",
+            "value": "100%"
+          }
+        },
+        "4e0ce38999fd431c971f095a11493445": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "FloatProgressModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "FloatProgressModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "ProgressView",
+            "bar_style": "success",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_c5dabb6aa34d4808bfe8a876a15130b4",
+            "max": 5148,
+            "min": 0,
+            "orientation": "horizontal",
+            "style": "IPY_MODEL_643947d50110467fab73bb2609e40467",
+            "value": 5148
+          }
+        },
+        "5daff87ee00f4024a4d0c36ec365656d": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "HTMLModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_dom_classes": [],
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "HTMLModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/controls",
+            "_view_module_version": "1.5.0",
+            "_view_name": "HTMLView",
+            "description": "",
+            "description_tooltip": null,
+            "layout": "IPY_MODEL_fffcb3453216492cbb954f130aeaba29",
+            "placeholder": "​",
+            "style": "IPY_MODEL_7d11aa9d9be3406581f64757a5550af7",
+            "value": " 5148/5148 [00:00&lt;00:00, 82853.79it/s]"
+          }
+        },
+        "89306a31ada645e887f280483191d061": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "83b00e40330e4a9bab89fce2d40aab7f": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "77b51cbd87e04728a058d46ee733c819": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        },
+        "c5dabb6aa34d4808bfe8a876a15130b4": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "643947d50110467fab73bb2609e40467": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "ProgressStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "ProgressStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "bar_color": null,
+            "description_width": ""
+          }
+        },
+        "fffcb3453216492cbb954f130aeaba29": {
+          "model_module": "@jupyter-widgets/base",
+          "model_name": "LayoutModel",
+          "model_module_version": "1.2.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/base",
+            "_model_module_version": "1.2.0",
+            "_model_name": "LayoutModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "LayoutView",
+            "align_content": null,
+            "align_items": null,
+            "align_self": null,
+            "border": null,
+            "bottom": null,
+            "display": null,
+            "flex": null,
+            "flex_flow": null,
+            "grid_area": null,
+            "grid_auto_columns": null,
+            "grid_auto_flow": null,
+            "grid_auto_rows": null,
+            "grid_column": null,
+            "grid_gap": null,
+            "grid_row": null,
+            "grid_template_areas": null,
+            "grid_template_columns": null,
+            "grid_template_rows": null,
+            "height": null,
+            "justify_content": null,
+            "justify_items": null,
+            "left": null,
+            "margin": null,
+            "max_height": null,
+            "max_width": null,
+            "min_height": null,
+            "min_width": null,
+            "object_fit": null,
+            "object_position": null,
+            "order": null,
+            "overflow": null,
+            "overflow_x": null,
+            "overflow_y": null,
+            "padding": null,
+            "right": null,
+            "top": null,
+            "visibility": null,
+            "width": null
+          }
+        },
+        "7d11aa9d9be3406581f64757a5550af7": {
+          "model_module": "@jupyter-widgets/controls",
+          "model_name": "DescriptionStyleModel",
+          "model_module_version": "1.5.0",
+          "state": {
+            "_model_module": "@jupyter-widgets/controls",
+            "_model_module_version": "1.5.0",
+            "_model_name": "DescriptionStyleModel",
+            "_view_count": null,
+            "_view_module": "@jupyter-widgets/base",
+            "_view_module_version": "1.2.0",
+            "_view_name": "StyleView",
+            "description_width": ""
+          }
+        }
+      }
+    }
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### WGAN\n",
-    "\n",
-    "* Modify snippets below and implement [Wasserstein GAN](https://arxiv.org/abs/1701.07875) with weight clipping. (2 points)\n",
-    "\n",
-    "* Replace  weight clipping with [gradient penalty](https://arxiv.org/pdf/1704.00028v3.pdf). (2 points)\n",
-    "\n",
-    "* Add labels into WGAN, performing [conditional generation](https://arxiv.org/pdf/1411.1784.pdf). (2 points) \n",
-    "\n",
-    "Write a report about experiments and results, add plots and visualizations."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import torch\n",
-    "import torch.nn as nn\n",
-    "import torch.nn.functional as F\n",
-    "import torch.optim as optim\n",
-    "from torch.utils.data import DataLoader, Dataset\n",
-    "\n",
-    "import torchvision\n",
-    "import matplotlib.pyplot as plt\n",
-    "import numpy as np\n",
-    "\n",
-    "from torch.autograd import Variable"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Creating config object (argparse workaround)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "class Config:\n",
-    "    pass\n",
-    "\n",
-    "config = Config()\n",
-    "config.mnist_path = None\n",
-    "config.batch_size = 16\n",
-    "config.num_workers = 3\n",
-    "config.num_epochs = 10\n",
-    "config.noise_size = 50\n",
-    "config.print_freq = 100\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Create dataloder"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "train = torchvision.datasets.FashionMNIST(\"fashion_mnist\", train=True, transform=torchvision.transforms.ToTensor(), download=True)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "dataloader = DataLoader(train, batch_size=16, shuffle=True)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "len(dataloader)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "for image, cat in dataloader:\n",
-    "    break"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [],
-   "source": [
-    "image.size()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Create generator and discriminator"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "class Generator(nn.Module):\n",
-    "    def __init__(self):\n",
-    "        super(Generator, self).__init__()\n",
-    "        self.model = nn.Sequential( \n",
-    "            nn.Linear(config.noise_size, 200),\n",
-    "            nn.ReLU(inplace=True),\n",
-    "            nn.Linear(200, 28*28),\n",
-    "            nn.Sigmoid())\n",
-    "        \n",
-    "    def forward(self, x):\n",
-    "        return self.model(x)\n",
-    "    \n",
-    "class Discriminator(nn.Module):\n",
-    "    def __init__(self):\n",
-    "        super(Discriminator, self).__init__()\n",
-    "        self.model = nn.Sequential(\n",
-    "            nn.Linear(28*28, 200),\n",
-    "            nn.ReLU(inplace=True),\n",
-    "            nn.Linear(200, 50),\n",
-    "            nn.ReLU(inplace=True),\n",
-    "            nn.Linear(50, 1), \n",
-    "            nn.Sigmoid())\n",
-    "    def forward(self, x):\n",
-    "        return self.model(x)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "generator = Generator()\n",
-    "discriminator = Discriminator()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Create optimizers and loss"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "optim_G = optim.Adam(params=generator.parameters(), lr=0.0001)\n",
-    "optim_D = optim.Adam(params=discriminator.parameters(), lr=0.0001)\n",
-    "\n",
-    "criterion = nn.BCELoss()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Create necessary variables"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "input = Variable(torch.FloatTensor(config.batch_size, 28*28))\n",
-    "noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size))\n",
-    "fixed_noise = Variable(torch.FloatTensor(config.batch_size, config.noise_size).normal_(0, 1))\n",
-    "label = Variable(torch.FloatTensor(config.batch_size))\n",
-    "real_label = 1\n",
-    "fake_label = 0"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### GAN"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [],
-   "source": [
-    "ERRD_x = np.zeros(config.num_epochs)\n",
-    "ERRD_z = np.zeros(config.num_epochs)\n",
-    "ERRG = np.zeros(config.num_epochs)\n",
-    "N = len(dataloader)\n",
-    "\n",
-    "for epoch in range(config.num_epochs):\n",
-    "    for iteration, (images, cat) in enumerate(dataloader):\n",
-    "        ####### \n",
-    "        # Discriminator stage: maximize log(D(x)) + log(1 - D(G(z))) \n",
-    "        #######\n",
-    "        discriminator.zero_grad()\n",
-    "        \n",
-    "        # real\n",
-    "        label.data.fill_(real_label)\n",
-    "        input_data = images.view(images.shape[0], -1)\n",
-    "        output = discriminator(input_data)\n",
-    "        errD_x = criterion(output, label)\n",
-    "        ERRD_x[epoch] += errD_x.item()\n",
-    "        errD_x.backward()\n",
-    "        \n",
-    "        # fake \n",
-    "        noise.data.normal_(0, 1)\n",
-    "        fake = generator(noise)\n",
-    "        label.data.fill_(fake_label)\n",
-    "        output = discriminator(fake.detach())\n",
-    "        errD_z = criterion(output, label)\n",
-    "        ERRD_z[epoch] += errD_z.item()\n",
-    "        errD_z.backward()\n",
-    "        \n",
-    "        optim_D.step()\n",
-    "        \n",
-    "        ####### \n",
-    "        # Generator stage: maximize log(D(G(x))\n",
-    "        #######\n",
-    "        generator.zero_grad()\n",
-    "        label.data.fill_(real_label)\n",
-    "        output = discriminator(fake)\n",
-    "        errG = criterion(output, label)\n",
-    "        ERRG[epoch] += errG.item()\n",
-    "        errG.backward()\n",
-    "        \n",
-    "        optim_G.step()\n",
-    "        \n",
-    "        if (iteration+1) % config.print_freq == 0:\n",
-    "            print('Epoch:{} Iter: {} errD_x: {:.2f} errD_z: {:.2f} errG: {:.2f}'.format(epoch+1,\n",
-    "                                                                                            iteration+1, \n",
-    "                                                                                            errD_x.item(),\n",
-    "                                                                                            errD_z.item(), \n",
-    "                                                                                            errG.item()))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "noise.data.normal_(0, 1)\n",
-    "fake = generator(noise)\n",
-    "\n",
-    "plt.figure(figsize=(6, 7))\n",
-    "for i in range(16):\n",
-    "    plt.subplot(4, 4, i + 1)\n",
-    "    plt.imshow(fake[i].detach().numpy().reshape(28, 28), cmap=plt.cm.Greys_r)\n",
-    "    plt.axis('off')"
-   ]
-  }
- ],
- "metadata": {
-  "anaconda-cloud": {},
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.8.11"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 1
-}
+  "nbformat": 4,
+  "nbformat_minor": 0
+}
\ No newline at end of file