image_inpainting_partial_convolution.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hardik1408/MADHAV-LABS/blob/main/task4_image_inpainting_partial_convolution.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zki_aml7ME1B"
      },
      "source": [
        "# Generative Inpainting using Partial Convolutions\n",
        "- I have implemented Image Inpainting using Partial Convolution.\n",
        "- Implementation has been done in Keras.\n",
        "\n",
        "\n",
        "\n",
        "Brief description of model:\n",
        "\n",
        "Partial Convolutional U-Net model with methods for building the model, compiling, and defining various loss functions. The class includes methods for training the model using a generator and making predictions. The model utilizes VGG16 for perceptual and style losses and includes custom loss functions such as hole loss, valid loss, and total variation loss. The generator function is used for training the model with a generator."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "isDft6P-9mcB"
      },
      "source": [
        "# **Installing and Importing Libraries**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "Zlj7Gg3gRhUN"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import torch\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "from torchvision import models\n",
        "import copy\n",
        "import cv2\n",
        "from PIL import Image\n",
        "import os\n",
        "import seaborn as sns\n",
        "import itertools"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2uHMFzQh9xm6"
      },
      "source": [
        "# **Loading the dataset**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "28SfqTG56D31",
        "outputId": "fcbd2507-2d18-438e-c45b-2861bf4ee239"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cuda:0\n"
          ]
        }
      ],
      "source": [
        "# defining the device\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "DzCxsyJ07kWz"
      },
      "outputs": [],
      "source": [
        "# Normalization and tensor conversion for both training and testing datasets using PyTorch's transforms\n",
        "transform_train = transforms.Compose([transforms.ToTensor(),\n",
        "                                      transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n",
        "])\n",
        "\n",
        "transform_test = transforms.Compose([transforms.ToTensor(),\n",
        "                                      transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),\n",
        "])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TdeaKI6mR33X",
        "outputId": "1cbe0ec1-cb20-4ca3-8325-a670c8afcfe0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "100%|██████████| 170498071/170498071 [00:02<00:00, 78881055.56it/s]\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Extracting ./data/cifar-10-python.tar.gz to ./data\n",
            "Files already downloaded and verified\n"
          ]
        }
      ],
      "source": [
        "# downloading the dataset\n",
        "trainset = torchvision.datasets.CIFAR10(root = './data', train = True, download = True, transform = transform_train)\n",
        "testset = torchvision.datasets.CIFAR10(root = './data', train = False, download = True, transform = transform_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "QM07cSpOyXeb"
      },
      "outputs": [],
      "source": [
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size=4, shuffle=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "_Z-L8_KByqhy"
      },
      "outputs": [],
      "source": [
        "# Fetching a batch of images and labels from a PyTorch DataLoader for training.\n",
        "dataiter = iter(trainloader)\n",
        "images, labels = next(dataiter)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "P5FCmLniy8e5"
      },
      "outputs": [],
      "source": [
        "img = images[0]\n",
        "npimg = img.numpy()\n",
        "npimg = np.transpose(npimg, (1,2,0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "Y5bxT9WgEfRO"
      },
      "outputs": [],
      "source": [
        "batch_size = 4"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "pPDwR6l3zPd4"
      },
      "outputs": [],
      "source": [
        "# Function to display images\n",
        "def imshow(img, title):\n",
        "  npimg = img.numpy() / 2 + 0.5\n",
        "  plt.figure(figsize = (batch_size, 1))\n",
        "  plt.axis('off')\n",
        "  plt.imshow(np.transpose(npimg, (1,2,0)))\n",
        "  plt.title(title)\n",
        "  plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 42,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 463
        },
        "id": "4LeTd00NzCB5",
        "outputId": "fd37df7b-302a-4dc4-fb12-0dbfa9d721fc"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 500x500 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.figure(figsize = (5,5))\n",
        "plt.imshow(npimg)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {
        "id": "2wIeW7rzzdfX"
      },
      "outputs": [],
      "source": [
        "# function to show images in batches\n",
        "def show_batch_images(dataloader):\n",
        "  images, labels = next(iter(dataloader))\n",
        "  img = torchvision.utils.make_grid(images)\n",
        "  imshow(img, title = \"example images\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 46,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "id": "H1K0UNFfEYAe",
        "outputId": "b3500c17-c206-4580-f31c-d0421a726cb3"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 400x100 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 400x100 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 400x100 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 400x100 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "for i in range(4):\n",
        "  show_batch_images(trainloader)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mzbiB9zYOq7D"
      },
      "source": [
        "# **Defining variuous classes**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "bkl2AzSogaxI"
      },
      "outputs": [],
      "source": [
        "# Custom PyTorch dataset class for CIFAR-10 images with an additional method to create masked images, generating random lines as masks.\n",
        "# The dataset returns input, mask, and output images for training.\n",
        "\n",
        "class Dataset(torchvision.datasets.CIFAR10):\n",
        "  def __init__(self, path, image_dims = (32, 32), n_channels = 3, train = True):\n",
        "    super().__init__(path, train, download = True)\n",
        "    self.n_classes = 10\n",
        "    self.image_dims = image_dims\n",
        "    self.n_channels = n_channels\n",
        "    self.resize = transforms.Resize(image_dims)\n",
        "    self.to_tensor = transforms.ToTensor()\n",
        "\n",
        "  def __getitem__(self, index):\n",
        "    image, label = super().__getitem__(index)\n",
        "\n",
        "    Y_output = self.resize(image)\n",
        "    X_input, X_mask = self.Create_Mask(Y_output)\n",
        "    return self.to_tensor(X_input), self.to_tensor(X_mask), self.to_tensor(Y_output)\n",
        "\n",
        "  def Create_Mask(self, image):\n",
        "    mask = np.full((self.image_dims[0], self.image_dims[1], self.n_channels), 255, dtype = np.uint8)\n",
        "\n",
        "    for _ in range(np.random.randint(1, 10)):\n",
        "      x1, x2 = np.random.randint(1, self.image_dims[0]), np.random.randint(1, self.image_dims[0])\n",
        "      y1, y2 = np.random.randint(1, self.image_dims[1]), np.random.randint(1, self.image_dims[1])\n",
        "      thickness = np.random.randint(1, 3)\n",
        "\n",
        "      cv2.line(mask, (x1, y1), (x2, y2), (1, 1, 1), thickness)\n",
        "\n",
        "    image = np.asarray(image)\n",
        "    masked_image = cv2.bitwise_and(image, mask)\n",
        "    return masked_image, mask"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "E9NYxC49O64Y"
      },
      "outputs": [],
      "source": [
        "# A class MaskGenerator that creates masks with random lines, circles, ellipses, or rectangles on an image for data augmentation, with parameters like scale ratio and shape randomness.\n",
        "class MaskGenerator():\n",
        "\n",
        "    def __init__ (self,X=None,scale_ratio = 0.03) :\n",
        "\n",
        "        self.height, self.width , self.channels = X.shape\n",
        "        self.mask = np.zeros((self.height,self.width,self.channels),dtype='uint8')\n",
        "        self.scale_ratio = scale_ratio\n",
        "        self.masker()\n",
        "\n",
        "\n",
        "    def masker(self,line=True,circle=True,ellipse=True,rectangle=False,polygon=False) :\n",
        "\n",
        "        scale = int((self.height+self.width)*self.scale_ratio)\n",
        "        if line :\n",
        "            for i in range(randint(1,15)):\n",
        "                x1,x2 = randint(1,self.width) ,randint(1,self.width)\n",
        "                y1,y2 = randint(1,self.height) ,randint(1,self.height)\n",
        "                cv2.line(self.mask,(x1,x2),(y1,y2),(1,1,1),randint(4,scale))\n",
        "        if circle :\n",
        "            for i in range(randint(1,15)):\n",
        "                cx,cy  = randint(1,self.width), randint(1,self.height)\n",
        "                radius = randint(30,50)\n",
        "                cv2.circle(self.mask, (cx,cy), radius, (1,1,1), randint(4,scale))\n",
        "        if ellipse :\n",
        "            for i in range(randint(1,15)):\n",
        "                x1, y1 = randint(1, self.width), randint(1, self.height)\n",
        "                s1, s2 = randint(1, self.width), randint(1, self.height)\n",
        "                a1, a2, a3 = randint(3, 180), randint(3, 180), randint(3, 180)\n",
        "                thickness = randint(20, scale)\n",
        "                cv2.ellipse(self.mask, (x1,y1), (s1,s2), a1, a2, a3,(1,1,1), thickness)\n",
        "\n",
        "        if rectangle :\n",
        "             for i in range(randint(1,15)):\n",
        "                x1,x2 = randint(1,self.width) ,randint(1,self.width)\n",
        "                y1,y2 = randint(1,self.height) ,randint(1,self.height)\n",
        "                cv2.rectangle(self.mask, (x1,y1), (x2,y2), (1,1,1), randint(4,20))\n",
        "\n",
        "        self.mask = 1 - self.mask"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7gSegnfrO8AI"
      },
      "outputs": [],
      "source": [
        "\n",
        "class PConvUNET(object):\n",
        "\n",
        "    def __init__(self,h,w,c,data_format='channels_last',weights='imagenet'):\n",
        "\n",
        "        self.h = h\n",
        "        self.w = w\n",
        "        self.c = c\n",
        "        self.data_format = data_format\n",
        "        self.mean = [0.485, 0.456, 0.406]\n",
        "        self.std = [0.229, 0.224, 0.225]\n",
        "        self.no_of_pixels = h*w*c\n",
        "        self.vgg_layers=[3,6,10]\n",
        "        self.vgg = self.vgg_model(weights)\n",
        "        self.model,self.masks = self.build_pconv_UNet()\n",
        "        self.model_compile(self.model,self.masks)\n",
        "\n",
        "\n",
        "    def vgg_model(self,weights):\n",
        "        '''A non-trainable vgg-model trained on imagenet\n",
        "           is used to extract features from higher level\n",
        "           layers namely : pool1,pool2,pool3 which will be\n",
        "           used for calculating perceptual and style losses'''\n",
        "        if weights == 'imagenet' :\n",
        "            vgg = VGG16(include_top=False,weights='imagenet')\n",
        "        else :\n",
        "            vgg = VGG16(weights=None, include_top=False)\n",
        "            vgg.load_weights(weights, by_name=True)\n",
        "        imgs = Input((self.h,self.w,self.c))\n",
        "        processed = Lambda(lambda x: (x-self.mean) / self.std)(imgs)\n",
        "        vgg.outputs = [vgg.layers[i].output for i in self.vgg_layers]\n",
        "        model = Model(inputs=imgs,outputs=vgg(processed))\n",
        "        model.trainable = False\n",
        "        model.compile(loss='mse',optimizer='adam')\n",
        "        return model\n",
        "\n",
        "    def build_pconv_UNet(self,train_bn=True):\n",
        "\n",
        "        inputs_img = Input((self.h, self.w, 3), name='inputs_img')\n",
        "        inputs_mask = Input((self.h, self.w, 3), name='inputs_mask')\n",
        "\n",
        "        def encoder_layer(imgs,masks,filters,kernel_size,bn=True):\n",
        "            I_out , Mask_out = Pconv(filters,kernel_size,strides=2,padding='same')([imgs,masks])\n",
        "            if bn :\n",
        "                I_out = BatchNormalization()(I_out,training=train_bn)\n",
        "            I_out = Activation('relu')(I_out)\n",
        "\n",
        "            return I_out,Mask_out\n",
        "\n",
        "        en_img1, en_mask1 = encoder_layer(inputs_img,inputs_mask,64,7,False)\n",
        "        en_img2, en_mask2 = encoder_layer(en_img1,en_mask1,128,5)\n",
        "        en_img3, en_mask3 = encoder_layer(en_img2,en_mask2,256,5)\n",
        "        en_img3, en_mask3 = encoder_layer(en_img2,en_mask2,256,5)\n",
        "        en_img3, en_mask3 = encoder_layer(en_img2,en_mask2,256,5)\n",
        "        en_img4, en_mask4 = encoder_layer(en_img3,en_mask3,512,3)\n",
        "        en_img5, en_mask5 = encoder_layer(en_img4,en_mask4,512,3)\n",
        "        en_img6, en_mask6 = encoder_layer(en_img5,en_mask5,512,3)\n",
        "        en_img7, en_mask7 = encoder_layer(en_img6,en_mask6,512,3)\n",
        "        en_img8, en_mask8 = encoder_layer(en_img7,en_mask7,512,3)\n",
        "\n",
        "\n",
        "        def decoder_layer(imgs,masks,en_img,en_mask,filters,kernel_size,bn=True,**kwargs):\n",
        "            I_up    = UpSampling2D(size=(2,2),data_format=self.data_format)(imgs)\n",
        "            mask_up = UpSampling2D(size=(2,2),data_format=self.data_format)(masks)\n",
        "            if self.data_format == 'channels_first' :\n",
        "                axis = 1\n",
        "            else :\n",
        "                axis = 3\n",
        "\n",
        "            I_up_concat    = Concatenate(axis)([I_up,en_img])\n",
        "            mask_up_concat = Concatenate(axis)([mask_up,en_mask])\n",
        "\n",
        "            I_out, Mask_out = Pconv(filters,kernel_size,padding='same')([I_up,mask_up])\n",
        "            if bn:\n",
        "                I_out = BatchNormalization()(I_out)\n",
        "            I_out = LeakyReLU(alpha=0.2)(I_out)\n",
        "            return I_out, Mask_out\n",
        "\n",
        "        dec_conv1, dec_mask1 = decoder_layer(en_img8,en_mask8,en_img7,en_mask7,512,3)\n",
        "        dec_conv2, dec_mask2 = decoder_layer(dec_conv1,dec_mask1,en_img6,en_mask6,512,3)\n",
        "        dec_conv3, dec_mask3 = decoder_layer(dec_conv2,dec_mask2,en_img5,en_mask5,512,3)\n",
        "        dec_conv4, dec_mask4 = decoder_layer(dec_conv3,dec_mask3,en_img4,en_mask4,512,3)\n",
        "        dec_conv5, dec_mask5 = decoder_layer(dec_conv4,dec_mask4,en_img3,en_mask3,256,3)\n",
        "        dec_conv6, dec_mask6 = decoder_layer(dec_conv5,dec_mask5,en_img2,en_mask2,128,3)\n",
        "        dec_conv7, dec_mask7 = decoder_layer(dec_conv6,dec_mask6,en_img1,en_mask1,64,3)\n",
        "        dec_conv8, dec_mask8 = decoder_layer(dec_conv7,dec_mask7,inputs_img,inputs_mask,3,3,bn=False)\n",
        "\n",
        "        outputs = Conv2D(3, 1, activation = 'sigmoid', name='outputs_img')(dec_conv8)\n",
        "        model = Model(inputs=[inputs_img, inputs_mask], outputs=outputs)\n",
        "\n",
        "        return model, inputs_mask\n",
        "\n",
        "\n",
        "    def model_compile(self,model,masks,lr=0.001):\n",
        "        model.compile(\n",
        "                     loss=self.loss_total(masks),\n",
        "                     optimizer= Adam(lr=lr),\n",
        "                     metrics=[self.PSNR]\n",
        "                     )\n",
        "\n",
        "\n",
        "    def loss_total(self,M):\n",
        "\n",
        "        def loss(I_out,I_gt):\n",
        "            '''I_comp is obtained by setting all the non hole pixels of I_out directly to ground truth'''\n",
        "            I_comp   = I_out*(1-M) + I_gt*M\n",
        "            vgg_out  = self.vgg(I_out)\n",
        "            vgg_comp = self.vgg(I_comp)\n",
        "            vgg_gt   = self.vgg(I_gt)\n",
        "\n",
        "            l1 = self.hole_loss(I_out,I_gt,M)\n",
        "            l2 = self.loss_valid(I_out,I_gt,M)\n",
        "            l3 = self.loss_perceptual(vgg_out,vgg_gt,vgg_comp)\n",
        "            l4 = self.style_loss(vgg_out,vgg_gt)\n",
        "            l5 = self.style_loss(vgg_comp,vgg_gt)\n",
        "            l6 = self.loss_tv(M,I_comp)\n",
        "\n",
        "            return l1 + 6*l2 + 0.05*l3 + 120*(l4+l5) + 0.1*l6\n",
        "        return loss\n",
        "\n",
        "\n",
        "    def hole_loss(self,I_out,I_gt,M):\n",
        "        '''Per pixel loss defined as the L1 distance between the ground truth and the output image\n",
        "           after multiplying by (1-M) where M is the corresponding mask'''\n",
        "        return self.l1((1-M)*I_out,(1-M)*I_gt)\n",
        "\n",
        "    def loss_valid(self,I_out,I_gt,M):\n",
        "        ''' loss_valid = L1 distance of M * (I_out-I_gt)'''\n",
        "        return self.l1(M*I_out,M*I_gt)\n",
        "\n",
        "    def loss_perceptual(self,vgg_out,vgg_gt,vgg_comp):\n",
        "        loss = 0\n",
        "        for i,j,k in zip(vgg_out,vgg_comp,vgg_gt) :\n",
        "            '''if K.ndim(i) == 4:\n",
        "                _,n1,n2,n3 = K.shape(i)\n",
        "            else :\n",
        "                n1,n2,n3 = K.shape(i)\n",
        "            n = n1*n2*n3'''\n",
        "            loss += self.l1(i,k) + self.l1(j,k)\n",
        "        return loss\n",
        "\n",
        "    def style_loss(self,output,vgg_gt):\n",
        "        loss = 0\n",
        "        for o, g in zip(output, vgg_gt):\n",
        "            loss += self.l1(self.gram_matrix(o), self.gram_matrix(g))\n",
        "        return loss\n",
        "\n",
        "    def loss_tv(self, mask, y_comp):\n",
        "        \"\"\"Total variation loss, used for smoothing the hole region, see. eq. 6\"\"\"\n",
        "\n",
        "        # Create dilated hole region using a 3x3 kernel of all 1s.\n",
        "        kernel = K.ones(shape=(3, 3, mask.shape[3], mask.shape[3]))\n",
        "        dilated_mask = K.conv2d(1-mask, kernel, data_format='channels_last', padding='same')\n",
        "\n",
        "        # Cast values to be [0., 1.], and compute dilated hole region of y_comp\n",
        "        dilated_mask = K.cast(K.greater(dilated_mask, 0), 'float32')\n",
        "        P = dilated_mask * y_comp\n",
        "\n",
        "        # Calculate total variation loss\n",
        "        a = self.l1(P[:,1:,:,:], P[:,:-1,:,:])\n",
        "        b = self.l1(P[:,:,1:,:], P[:,:,:-1,:])\n",
        "        return a+b\n",
        "\n",
        "    def model_summary(self):\n",
        "        return self.model.summary()\n",
        "\n",
        "\n",
        "    def generator(self,generator,*args,**kwargs):\n",
        "        self.model.fit_generator(generator,*args,**kwargs)\n",
        "\n",
        "    ##############################################################################\n",
        "    '''Gram-matrix function taken from github.\n",
        "       It is required for the style loss terms'''\n",
        "\n",
        "    def gram_matrix(self,x, norm_by_channels=False):\n",
        "        \"\"\"Calculate gram matrix used in style loss\"\"\"\n",
        "\n",
        "        # Assertions on input\n",
        "        assert K.ndim(x) == 4, 'Input tensor should be a 4d (B, H, W, C) tensor'\n",
        "        assert K.image_data_format() == 'channels_last', \"Please use channels-last format\"\n",
        "\n",
        "        # Permute channels and get resulting shape\n",
        "        x = K.permute_dimensions(x, (0, 3, 1, 2))\n",
        "        shape = K.shape(x)\n",
        "        B, C, H, W = shape[0], shape[1], shape[2], shape[3]\n",
        "\n",
        "        # Reshape x and do batch dot product\n",
        "        features = K.reshape(x, K.stack([B, C, H*W]))\n",
        "        gram = K.batch_dot(features, features, axes=2)\n",
        "\n",
        "        # Normalize with channels, height and width\n",
        "        gram = gram /  K.cast(C * H * W, x.dtype)\n",
        "\n",
        "        return gram\n",
        "    ################################################################################################\n",
        "\n",
        "\n",
        "    def l1(self,y_true, y_pred):\n",
        "        \"\"\"Calculate the L1 loss used in all loss calculations\"\"\"\n",
        "        if K.ndim(y_true) == 4:\n",
        "            return K.mean(K.abs(y_pred - y_true), axis=[1,2,3])\n",
        "        elif K.ndim(y_true) == 3:\n",
        "            return K.mean(K.abs(y_pred - y_true), axis=[1,2])\n",
        "        else:\n",
        "            raise NotImplementedError(\"Calculating L1 loss on 1D tensors? should not occur for this network\")\n",
        "\n",
        "    def PSNR(self,y_true, y_pred):\n",
        "        \"\"\"\n",
        "        PSNR is Peek Signal to Noise Ratio, see https://en.wikipedia.org/wiki/Peak_signal-to-noise_ratio\n",
        "        The equation is:\n",
        "        PSNR = 20 * log10(MAX_I) - 10 * log10(MSE)\n",
        "\n",
        "        Our input is scaled with be within the range -2.11 to 2.64 (imagenet value scaling). We use the difference between these\n",
        "        two values (4.75) as MAX_I\n",
        "        \"\"\"\n",
        "        #return 20 * K.log(4.75) / K.log(10.0) - 10.0 * K.log(K.mean(K.square(y_pred - y_true))) / K.log(10.0)\n",
        "        return - 10.0 * K.log(K.mean(K.square(y_pred - y_true))) / K.log(10.0)\n",
        "\n",
        "\n",
        "    def predict(self,test,**kwargs):\n",
        "        return self.model.predict(test,**kwargs)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vkuao1ik8s8F"
      },
      "outputs": [],
      "source": [
        "class PartialConv(nn.Module):\n",
        "    def __init__(self, in_filters, out_filters, kernel_size, strides=1, padding=0, dilation=1, groups=1, bias=True):\n",
        "        super(PartialConv, self).__init__()\n",
        "\n",
        "        padding = (kernel_size[0] // 2, kernel_size[1] // 2)\n",
        "        self.input_conv = nn.Conv2d(in_filters, out_filters, kernel_size, strides, padding, bias=bias)\n",
        "        self.mask_conv = nn.Conv2d(in_filters, out_filters, kernel_size, strides, padding, bias=False)\n",
        "\n",
        "        self.window_size = kernel_size[0] * kernel_size[1]\n",
        "\n",
        "        nn.init.constant_(self.mask_conv.weight, 1.0)\n",
        "\n",
        "        for param in self.mask_conv.parameters():\n",
        "            param.requires_grad = False\n",
        "\n",
        "    def forward(self, input, mask):\n",
        "        output = self.input_conv(input * mask)\n",
        "        if self.input_conv.bias is not None:\n",
        "            output_bias = self.input_conv.bias.view(1, -1, 1, 1).expand_as(output)\n",
        "        else:\n",
        "            output_bias = torch.zeros_like(output)\n",
        "\n",
        "        with torch.no_grad():\n",
        "            output_mask = self.mask_conv(mask)\n",
        "\n",
        "        mask_ratio = self.window_size / (output_mask + 1e-8)\n",
        "\n",
        "        output_mask = torch.clamp(output_mask, 0, 1)\n",
        "\n",
        "        mask_ratio = mask_ratio * output_mask\n",
        "\n",
        "        output = output * mask_ratio\n",
        "\n",
        "        output = output + output_bias\n",
        "\n",
        "        return output, output_mask"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OEaXFnPY8tQd"
      },
      "outputs": [],
      "source": [
        "# a class defining functions for the encoder layer\n",
        "class EncoderLayer(nn.Module):\n",
        "    def __init__(self, in_filters, out_filters):\n",
        "        super(EncoderLayer, self).__init__()\n",
        "        self.p_conv_1 = PartialConv(in_filters, out_filters, (3, 3), strides=1)\n",
        "        self.act_1 = nn.ReLU()\n",
        "        self.p_conv_2 = PartialConv(out_filters, out_filters, (3, 3), strides=2)\n",
        "        self.act_2 = nn.ReLU()\n",
        "\n",
        "    def forward(self, inputs, masks):\n",
        "        conv_out_1, mask_out_1 = self.p_conv_1(inputs, masks)\n",
        "        conv_out_1 = self.act_1(conv_out_1)\n",
        "\n",
        "        conv_out_2, mask_out_2 = self.p_conv_2(conv_out_1, mask_out_1)\n",
        "        conv_out_2 = self.act_2(conv_out_2)\n",
        "\n",
        "        return conv_out_1, mask_out_1, conv_out_2, mask_out_2"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YD3-pJig8tc0"
      },
      "outputs": [],
      "source": [
        "# a class defining functions for the decoder layer\n",
        "class DecoderLayer(nn.Module):\n",
        "    def __init__(self, in_filters_1, in_filters_2):\n",
        "        super(DecoderLayer, self).__init__()\n",
        "\n",
        "        self.up_image = nn.Upsample(scale_factor=(2, 2))\n",
        "        self.up_mask = nn.Upsample(scale_factor=(2, 2))\n",
        "\n",
        "        self.p_conv_1 = PartialConv(in_filters_1 * 2, in_filters_1, (3, 3))\n",
        "        self.act_1 = nn.LeakyReLU(negative_slope=0.2)\n",
        "        self.p_conv_2 = PartialConv(in_filters_1, in_filters_2, (3, 3))\n",
        "        self.act_2 = nn.LeakyReLU(negative_slope=0.2)\n",
        "\n",
        "    def forward(self, in_image, in_mask, share_image, share_mask):\n",
        "        up_out = self.up_image(in_image)\n",
        "        mask_out = self.up_mask(in_mask)\n",
        "\n",
        "        concat_image = torch.cat([share_image, up_out], axis=1)\n",
        "        concat_mask = torch.cat([share_mask, mask_out], axis=1)\n",
        "\n",
        "        conv_out_1, mask_out_1 = self.p_conv_1(concat_image, concat_mask)\n",
        "        conv_out_1 = self.act_1(conv_out_1)\n",
        "\n",
        "        conv_out_2, mask_out_2 = self.p_conv_2(conv_out_1, mask_out_1)\n",
        "        conv_out_2 = self.act_2(conv_out_2)\n",
        "\n",
        "        return conv_out_1, mask_out_1, conv_out_2, mask_out_2"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_cb8Dahv8trH"
      },
      "outputs": [],
      "source": [
        "# a class defining functions for the inpainting module\n",
        "class InpaintingModel(nn.Module):\n",
        "    def __init__(self, input_dims=(32, 32), num_channels=3):\n",
        "        super(InpaintingModel, self).__init__()\n",
        "\n",
        "        self.encoder_1 = EncoderLayer(num_channels, 32)\n",
        "        self.encoder_2 = EncoderLayer(32, 64)\n",
        "        self.encoder_3 = EncoderLayer(64, 128)\n",
        "        self.encoder_4 = EncoderLayer(128, 256)\n",
        "\n",
        "        self.decoder_1 = DecoderLayer(256, 128)\n",
        "        self.decoder_2 = DecoderLayer(128, 64)\n",
        "        self.decoder_3 = DecoderLayer(64, 32)\n",
        "        self.decoder_4 = DecoderLayer(32, 3)\n",
        "\n",
        "        self.output_layer = nn.Sequential(\n",
        "            nn.Conv2d(3, 3, (3, 3), padding=1),\n",
        "            nn.Sigmoid()\n",
        "        )\n",
        "\n",
        "    def forward(self, inputs, masks):\n",
        "        conv_out_1, mask_out_1, conv_out_2, mask_out_2 = self.encoder_1(inputs, masks)\n",
        "        conv_out_3, mask_out_3, conv_out_4, mask_out_4 = self.encoder_2(conv_out_2, mask_out_2)\n",
        "        conv_out_5, mask_out_5, conv_out_6, mask_out_6 = self.encoder_3(conv_out_4, mask_out_4)\n",
        "        conv_out_7, mask_out_7, conv_out_8, mask_out_8 = self.encoder_4(conv_out_6, mask_out_6)\n",
        "\n",
        "        conv_out_9, mask_out_9, conv_out_10, mask_out_10 = self.decoder_1(conv_out_8, mask_out_8, conv_out_7, mask_out_7)\n",
        "        conv_out_11, mask_out_11, conv_out_12, mask_out_12 = self.decoder_2(conv_out_6, mask_out_6, conv_out_5, mask_out_5)\n",
        "        conv_out_13, mask_out_13, conv_out_14, mask_out_14 = self.decoder_3(conv_out_4, mask_out_4, conv_out_3, mask_out_3)\n",
        "        conv_out_15, mask_out_15, conv_out_16, mask_out_16 = self.decoder_4(conv_out_2, mask_out_2, conv_out_1, mask_out_1)\n",
        "\n",
        "        outputs = self.output_layer(conv_out_16)\n",
        "        return outputs"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MOGPoiMRqFE0"
      },
      "source": [
        "# **Data Pre Proccessing**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "kOW7NfQ8gakJ"
      },
      "outputs": [],
      "source": [
        "image_dims = (32, 32)\n",
        "batch_size = 4"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "jnL93nw8qLuD"
      },
      "outputs": [],
      "source": [
        "def imshow_masked(img):\n",
        "  npimg = img.numpy()\n",
        "  plt.figure(figsize = (3, 3))\n",
        "  plt.axis('off')\n",
        "  plt.imshow(np.transpose(npimg, (1,2,0)))\n",
        "  plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oGAS2nwwiSzc",
        "outputId": "fd74b6b5-2598-4a97-fd48-5a8654d4f4db"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Files already downloaded and verified\n",
            "Files already downloaded and verified\n"
          ]
        }
      ],
      "source": [
        "Trainset = Dataset('./data', image_dims)\n",
        "Testset = Dataset('./data', image_dims, train=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "nGM_0vCoidvi"
      },
      "outputs": [],
      "source": [
        "Trainloader = torch.utils.data.DataLoader(Trainset, batch_size = batch_size, shuffle = True)\n",
        "Testloader = torch.utils.data.DataLoader(Testset, batch_size = batch_size, shuffle = False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "n_sDI9TdiuaS"
      },
      "outputs": [],
      "source": [
        "Train_dataiter = iter(Trainloader)\n",
        "Test_dataiter = iter(Testloader)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2EJeMcQmiyUx",
        "outputId": "eafc6b25-6ffa-45eb-ce78-3be68b5f60ef"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "torch.Size([4, 3, 32, 32]) torch.Size([4, 3, 32, 32]) torch.Size([4, 3, 32, 32])\n"
          ]
        }
      ],
      "source": [
        "X_in, X_mask, Y_out = next(Train_dataiter)\n",
        "print(X_in.shape, X_mask.shape, Y_out.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        },
        "id": "HcyxPoNKkZad",
        "outputId": "2bcd91e5-2c44-46ad-f099-2e7f4039910a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Masked Image\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 300x300 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Masking Pattern\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 300x300 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Original Image\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 300x300 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# printing example images\n",
        "print('Masked Image')\n",
        "imshow_masked(torchvision.utils.make_grid(X_in))\n",
        "print('Masking Pattern')\n",
        "imshow_masked(torchvision.utils.make_grid(X_mask))\n",
        "print('Original Image')\n",
        "imshow_masked(torchvision.utils.make_grid(Y_out))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hi-Fgw-D9zCu"
      },
      "source": [
        "# **Building the model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "G1dJ9I6qmYiW"
      },
      "outputs": [],
      "source": [
        "#  computation of the Jaccard coefficient\n",
        "def jaccard_coeff(dataloader):\n",
        "  for (X_input, X_mask, Y_output) in dataloader:\n",
        "    X_input, X_mask, Y_output = X_input.to(device), X_mask.to(device), Y_output.to(device)\n",
        "    outputs = model(X_input, X_mask)\n",
        "    eps = 1e-15\n",
        "    outputs = outputs.view(-1)\n",
        "    Y_output = Y_output.view(-1)\n",
        "    intersection = (outputs * Y_output).sum()\n",
        "    union = (outputs.sum() + Y_output.sum()) - intersection\n",
        "  return (intersection / (union + eps))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "FgswJGqwcM82"
      },
      "outputs": [],
      "source": [
        "# defining the models\n",
        "model = InpaintingModel()\n",
        "model = model.to(device)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "Xm7FbZUsrM-P"
      },
      "outputs": [],
      "source": [
        "# defining the loss function and optimizer\n",
        "loss_fn = nn.MSELoss()\n",
        "opt = optim.Adam(model.parameters())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "7txrvGRpymsM",
        "outputId": "5ac6ef8c-071b-4143-878c-4b3659e63d34"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Min loss 0.08\n",
            "Iteration: 0/12500, Loss: 0.08\n",
            "Min loss 0.07\n",
            "Min loss 0.04\n",
            "Min loss 0.03\n",
            "Iteration: 100/12500, Loss: 0.08\n",
            "Min loss 0.03\n",
            "Min loss 0.03\n",
            "Min loss 0.03\n",
            "Min loss 0.03\n",
            "Min loss 0.02\n",
            "Min loss 0.02\n",
            "Min loss 0.02\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Iteration: 200/12500, Loss: 0.02\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Iteration: 300/12500, Loss: 0.01\n",
            "Min loss 0.01\n",
            "Min loss 0.01\n",
            "Iteration: 400/12500, Loss: 0.01\n",
            "Iteration: 500/12500, Loss: 0.01\n",
            "Min loss 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 600/12500, Loss: 0.01\n",
            "Iteration: 700/12500, Loss: 0.01\n",
            "Iteration: 800/12500, Loss: 0.01\n",
            "Iteration: 900/12500, Loss: 0.01\n",
            "Iteration: 1000/12500, Loss: 0.01\n",
            "Iteration: 1100/12500, Loss: 0.01\n",
            "Iteration: 1200/12500, Loss: 0.02\n",
            "Iteration: 1300/12500, Loss: 0.02\n",
            "Iteration: 1400/12500, Loss: 0.01\n",
            "Iteration: 1500/12500, Loss: 0.02\n",
            "Iteration: 1600/12500, Loss: 0.01\n",
            "Iteration: 1700/12500, Loss: 0.01\n",
            "Iteration: 1800/12500, Loss: 0.02\n",
            "Iteration: 1900/12500, Loss: 0.02\n",
            "Iteration: 2000/12500, Loss: 0.01\n",
            "Iteration: 2100/12500, Loss: 0.02\n",
            "Iteration: 2200/12500, Loss: 0.02\n",
            "Iteration: 2300/12500, Loss: 0.01\n",
            "Iteration: 2400/12500, Loss: 0.01\n",
            "Iteration: 2500/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 2600/12500, Loss: 0.01\n",
            "Iteration: 2700/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 2800/12500, Loss: 0.01\n",
            "Iteration: 2900/12500, Loss: 0.01\n",
            "Iteration: 3000/12500, Loss: 0.01\n",
            "Iteration: 3100/12500, Loss: 0.01\n",
            "Iteration: 3200/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 3300/12500, Loss: 0.01\n",
            "Iteration: 3400/12500, Loss: 0.01\n",
            "Iteration: 3500/12500, Loss: 0.01\n",
            "Iteration: 3600/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 3700/12500, Loss: 0.01\n",
            "Iteration: 3800/12500, Loss: 0.01\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.01\n",
            "Iteration: 4100/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 4200/12500, Loss: 0.01\n",
            "Iteration: 4300/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 4400/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 4500/12500, Loss: 0.01\n",
            "Iteration: 4600/12500, Loss: 0.01\n",
            "Iteration: 4700/12500, Loss: 0.01\n",
            "Iteration: 4800/12500, Loss: 0.01\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.01\n",
            "Iteration: 5100/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 5200/12500, Loss: 0.01\n",
            "Iteration: 5300/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 5400/12500, Loss: 0.01\n",
            "Iteration: 5500/12500, Loss: 0.01\n",
            "Iteration: 5600/12500, Loss: 0.01\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.01\n",
            "Iteration: 5900/12500, Loss: 0.01\n",
            "Iteration: 6000/12500, Loss: 0.01\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.01\n",
            "Iteration: 6300/12500, Loss: 0.01\n",
            "Iteration: 6400/12500, Loss: 0.01\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.01\n",
            "Iteration: 6900/12500, Loss: 0.01\n",
            "Iteration: 7000/12500, Loss: 0.01\n",
            "Iteration: 7100/12500, Loss: 0.01\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 7400/12500, Loss: 0.01\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.01\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.01\n",
            "Iteration: 8400/12500, Loss: 0.01\n",
            "Iteration: 8500/12500, Loss: 0.01\n",
            "Iteration: 8600/12500, Loss: 0.01\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.01\n",
            "Iteration: 9100/12500, Loss: 0.01\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.01\n",
            "Iteration: 9400/12500, Loss: 0.01\n",
            "Min loss 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.01\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.01\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.01\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.01\n",
            "Iteration: 11400/12500, Loss: 0.01\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.01\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.01\n",
            "Iteration: 12100/12500, Loss: 0.01\n",
            "Iteration: 12200/12500, Loss: 0.01\n",
            "Iteration: 12300/12500, Loss: 0.01\n",
            "Iteration: 12400/12500, Loss: 0.01\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.01\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.01\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.01\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.01\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.01\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.01\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.01\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.01\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 8800/12500, Loss: 0.01\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.01\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.01\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.01\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.01\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.01\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 1200/12500, Loss: 0.01\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.01\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.01\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.01\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.01\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.01\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.01\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.01\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.01\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.01\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.01\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.01\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.01\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.01\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.01\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.01\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n",
            "Iteration: 0/12500, Loss: 0.00\n",
            "Iteration: 100/12500, Loss: 0.00\n",
            "Iteration: 200/12500, Loss: 0.00\n",
            "Iteration: 300/12500, Loss: 0.00\n",
            "Iteration: 400/12500, Loss: 0.00\n",
            "Iteration: 500/12500, Loss: 0.00\n",
            "Iteration: 600/12500, Loss: 0.00\n",
            "Iteration: 700/12500, Loss: 0.00\n",
            "Iteration: 800/12500, Loss: 0.00\n",
            "Iteration: 900/12500, Loss: 0.00\n",
            "Iteration: 1000/12500, Loss: 0.00\n",
            "Iteration: 1100/12500, Loss: 0.00\n",
            "Iteration: 1200/12500, Loss: 0.00\n",
            "Iteration: 1300/12500, Loss: 0.00\n",
            "Iteration: 1400/12500, Loss: 0.00\n",
            "Iteration: 1500/12500, Loss: 0.00\n",
            "Iteration: 1600/12500, Loss: 0.00\n",
            "Iteration: 1700/12500, Loss: 0.00\n",
            "Iteration: 1800/12500, Loss: 0.00\n",
            "Iteration: 1900/12500, Loss: 0.00\n",
            "Iteration: 2000/12500, Loss: 0.00\n",
            "Iteration: 2100/12500, Loss: 0.00\n",
            "Iteration: 2200/12500, Loss: 0.00\n",
            "Iteration: 2300/12500, Loss: 0.00\n",
            "Iteration: 2400/12500, Loss: 0.00\n",
            "Iteration: 2500/12500, Loss: 0.00\n",
            "Iteration: 2600/12500, Loss: 0.00\n",
            "Iteration: 2700/12500, Loss: 0.00\n",
            "Iteration: 2800/12500, Loss: 0.00\n",
            "Iteration: 2900/12500, Loss: 0.00\n",
            "Iteration: 3000/12500, Loss: 0.00\n",
            "Iteration: 3100/12500, Loss: 0.00\n",
            "Iteration: 3200/12500, Loss: 0.00\n",
            "Iteration: 3300/12500, Loss: 0.00\n",
            "Iteration: 3400/12500, Loss: 0.00\n",
            "Iteration: 3500/12500, Loss: 0.00\n",
            "Iteration: 3600/12500, Loss: 0.00\n",
            "Iteration: 3700/12500, Loss: 0.00\n",
            "Iteration: 3800/12500, Loss: 0.00\n",
            "Iteration: 3900/12500, Loss: 0.00\n",
            "Iteration: 4000/12500, Loss: 0.00\n",
            "Iteration: 4100/12500, Loss: 0.00\n",
            "Iteration: 4200/12500, Loss: 0.00\n",
            "Iteration: 4300/12500, Loss: 0.00\n",
            "Iteration: 4400/12500, Loss: 0.00\n",
            "Iteration: 4500/12500, Loss: 0.00\n",
            "Iteration: 4600/12500, Loss: 0.00\n",
            "Iteration: 4700/12500, Loss: 0.00\n",
            "Iteration: 4800/12500, Loss: 0.00\n",
            "Iteration: 4900/12500, Loss: 0.00\n",
            "Iteration: 5000/12500, Loss: 0.00\n",
            "Iteration: 5100/12500, Loss: 0.00\n",
            "Iteration: 5200/12500, Loss: 0.00\n",
            "Iteration: 5300/12500, Loss: 0.00\n",
            "Iteration: 5400/12500, Loss: 0.00\n",
            "Iteration: 5500/12500, Loss: 0.00\n",
            "Iteration: 5600/12500, Loss: 0.00\n",
            "Iteration: 5700/12500, Loss: 0.00\n",
            "Iteration: 5800/12500, Loss: 0.00\n",
            "Iteration: 5900/12500, Loss: 0.00\n",
            "Iteration: 6000/12500, Loss: 0.00\n",
            "Iteration: 6100/12500, Loss: 0.00\n",
            "Iteration: 6200/12500, Loss: 0.00\n",
            "Iteration: 6300/12500, Loss: 0.00\n",
            "Iteration: 6400/12500, Loss: 0.00\n",
            "Iteration: 6500/12500, Loss: 0.00\n",
            "Iteration: 6600/12500, Loss: 0.00\n",
            "Iteration: 6700/12500, Loss: 0.00\n",
            "Iteration: 6800/12500, Loss: 0.00\n",
            "Iteration: 6900/12500, Loss: 0.00\n",
            "Iteration: 7000/12500, Loss: 0.00\n",
            "Iteration: 7100/12500, Loss: 0.00\n",
            "Iteration: 7200/12500, Loss: 0.00\n",
            "Iteration: 7300/12500, Loss: 0.00\n",
            "Iteration: 7400/12500, Loss: 0.00\n",
            "Iteration: 7500/12500, Loss: 0.00\n",
            "Iteration: 7600/12500, Loss: 0.00\n",
            "Iteration: 7700/12500, Loss: 0.00\n",
            "Iteration: 7800/12500, Loss: 0.00\n",
            "Iteration: 7900/12500, Loss: 0.00\n",
            "Iteration: 8000/12500, Loss: 0.00\n",
            "Iteration: 8100/12500, Loss: 0.00\n",
            "Iteration: 8200/12500, Loss: 0.00\n",
            "Iteration: 8300/12500, Loss: 0.00\n",
            "Iteration: 8400/12500, Loss: 0.00\n",
            "Iteration: 8500/12500, Loss: 0.00\n",
            "Iteration: 8600/12500, Loss: 0.00\n",
            "Iteration: 8700/12500, Loss: 0.00\n",
            "Iteration: 8800/12500, Loss: 0.00\n",
            "Iteration: 8900/12500, Loss: 0.00\n",
            "Iteration: 9000/12500, Loss: 0.00\n",
            "Iteration: 9100/12500, Loss: 0.00\n",
            "Iteration: 9200/12500, Loss: 0.00\n",
            "Iteration: 9300/12500, Loss: 0.00\n",
            "Iteration: 9400/12500, Loss: 0.00\n",
            "Iteration: 9500/12500, Loss: 0.00\n",
            "Iteration: 9600/12500, Loss: 0.00\n",
            "Iteration: 9700/12500, Loss: 0.00\n",
            "Iteration: 9800/12500, Loss: 0.00\n",
            "Iteration: 9900/12500, Loss: 0.00\n",
            "Iteration: 10000/12500, Loss: 0.00\n",
            "Iteration: 10100/12500, Loss: 0.00\n",
            "Iteration: 10200/12500, Loss: 0.00\n",
            "Iteration: 10300/12500, Loss: 0.00\n",
            "Iteration: 10400/12500, Loss: 0.00\n",
            "Iteration: 10500/12500, Loss: 0.00\n",
            "Iteration: 10600/12500, Loss: 0.00\n",
            "Min loss 0.00\n",
            "Iteration: 10700/12500, Loss: 0.00\n",
            "Iteration: 10800/12500, Loss: 0.00\n",
            "Iteration: 10900/12500, Loss: 0.00\n",
            "Iteration: 11000/12500, Loss: 0.00\n",
            "Iteration: 11100/12500, Loss: 0.00\n",
            "Iteration: 11200/12500, Loss: 0.00\n",
            "Iteration: 11300/12500, Loss: 0.00\n",
            "Iteration: 11400/12500, Loss: 0.00\n",
            "Iteration: 11500/12500, Loss: 0.00\n",
            "Iteration: 11600/12500, Loss: 0.00\n",
            "Iteration: 11700/12500, Loss: 0.00\n",
            "Iteration: 11800/12500, Loss: 0.00\n",
            "Iteration: 11900/12500, Loss: 0.00\n",
            "Iteration: 12000/12500, Loss: 0.00\n",
            "Iteration: 12100/12500, Loss: 0.00\n",
            "Iteration: 12200/12500, Loss: 0.00\n",
            "Iteration: 12300/12500, Loss: 0.00\n",
            "Iteration: 12400/12500, Loss: 0.00\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 22min 46s, sys: 1min 47s, total: 24min 34s\n",
            "Wall time: 26min 4s\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "loss_epoch_arr = []\n",
        "max_epochs = 10\n",
        "\n",
        "min_loss = 1000\n",
        "\n",
        "n_iters = np.ceil(50000/batch_size)\n",
        "for epoch in range(max_epochs):\n",
        "    model.train()\n",
        "    for i, (X_input, X_mask, Y_output) in enumerate(Trainloader, 0):\n",
        "\n",
        "        X_input, X_mask, Y_output = X_input.to(device), X_mask.to(device), Y_output.to(device)\n",
        "\n",
        "        opt.zero_grad()\n",
        "\n",
        "        outputs = model(X_input, X_mask)\n",
        "        loss = loss_fn(Y_output, outputs)\n",
        "        loss.backward()\n",
        "\n",
        "        opt.step()\n",
        "\n",
        "        if min_loss > loss.item():\n",
        "            min_loss = loss.item()\n",
        "            best_model = copy.deepcopy(model.state_dict())\n",
        "            print('Min loss %0.2f' % min_loss)\n",
        "\n",
        "        if i % 100 == 0:\n",
        "            print('Iteration: %d/%d, Loss: %0.002f' % (i, n_iters, loss.item()))\n",
        "\n",
        "        del X_input, X_mask, Y_output\n",
        "        torch.cuda.empty_cache()\n",
        "\n",
        "    loss_epoch_arr.append(loss.item())\n",
        "\n",
        "plt.plot(loss_epoch_arr)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eQl0sR58Rg0z"
      },
      "source": [
        "# **Performance of the model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eK2kdLdwypS0",
        "outputId": "df0c86fc-b4de-4754-b417-7dc666a5668f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "execution_count": 32,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "model.load_state_dict(best_model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "62pmuzR43EPa",
        "outputId": "02547332-5371-45ae-811d-63f19a38da94"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "InpaintingModel(\n",
            "  (encoder_1): EncoderLayer(\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): ReLU()\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): ReLU()\n",
            "  )\n",
            "  (encoder_2): EncoderLayer(\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): ReLU()\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): ReLU()\n",
            "  )\n",
            "  (encoder_3): EncoderLayer(\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): ReLU()\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): ReLU()\n",
            "  )\n",
            "  (encoder_4): EncoderLayer(\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): ReLU()\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): ReLU()\n",
            "  )\n",
            "  (decoder_1): DecoderLayer(\n",
            "    (up_image): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (up_mask): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(512, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(512, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): LeakyReLU(negative_slope=0.2)\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): LeakyReLU(negative_slope=0.2)\n",
            "  )\n",
            "  (decoder_2): DecoderLayer(\n",
            "    (up_image): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (up_mask): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(256, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): LeakyReLU(negative_slope=0.2)\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): LeakyReLU(negative_slope=0.2)\n",
            "  )\n",
            "  (decoder_3): DecoderLayer(\n",
            "    (up_image): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (up_mask): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): LeakyReLU(negative_slope=0.2)\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): LeakyReLU(negative_slope=0.2)\n",
            "  )\n",
            "  (decoder_4): DecoderLayer(\n",
            "    (up_image): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (up_mask): Upsample(scale_factor=(2.0, 2.0), mode='nearest')\n",
            "    (p_conv_1): PartialConv(\n",
            "      (input_conv): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_1): LeakyReLU(negative_slope=0.2)\n",
            "    (p_conv_2): PartialConv(\n",
            "      (input_conv): Conv2d(32, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (mask_conv): Conv2d(32, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "    )\n",
            "    (act_2): LeakyReLU(negative_slope=0.2)\n",
            "  )\n",
            "  (output_layer): Sequential(\n",
            "    (0): Conv2d(3, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (1): Sigmoid()\n",
            "  )\n",
            ")\n"
          ]
        }
      ],
      "source": [
        "print(model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "id": "lXuQ25kG3PSQ"
      },
      "outputs": [],
      "source": [
        "weight_layer0 = list(model.parameters())[0].cpu()\n",
        "weight_layer0 = weight_layer0.data.numpy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "Nxo7si1PRWy3",
        "outputId": "d54db1a8-810c-47dd-8801-54e1be3cbf80"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#histogram to show the distribution of values\n",
        "plt.hist(weight_layer0.ravel(), bins=50, density=True, alpha=0.7, color='b')\n",
        "plt.xlabel('Weight Values')\n",
        "plt.ylabel('Frequency')\n",
        "plt.title('Weight Distribution - Layer 0')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "z-Opjh4oR_en",
        "outputId": "3516de01-747b-44c0-a6d5-e646cbda4631"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.violinplot(y=weight_layer0.ravel())\n",
        "plt.xlabel('Layer 0 Weights')\n",
        "plt.title('Violin Plot of Layer 0 Weights')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 38,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "cHt4oPKcSE8_",
        "outputId": "986937f8-02bb-4ea6-a4d1-ee2dc9981b65"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.scatter(range(len(weight_layer0.ravel())), weight_layer0.ravel(), alpha=0.5)\n",
        "plt.xlabel('Index')\n",
        "plt.ylabel('Weight Values')\n",
        "plt.title('Scatter Plot of Layer 0 Weights')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 652
        },
        "id": "ZCgyFKFx371o",
        "outputId": "e1fff306-562c-4a99-d660-d8401f912cbd"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-36-dfd95a8f0242>:1: UserWarning: \n",
            "\n",
            "`distplot` is a deprecated function and will be removed in seaborn v0.14.0.\n",
            "\n",
            "Please adapt your code to use either `displot` (a figure-level function with\n",
            "similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "\n",
            "For a guide to updating your code to use the new functions, please see\n",
            "https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751\n",
            "\n",
            "  sns.distplot(weight_layer0.ravel())\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "[]"
            ]
          },
          "execution_count": 36,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.distplot(weight_layer0.ravel())\n",
        "plt.plot()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 833
        },
        "id": "zbdagcLJFBCU",
        "outputId": "a4b77f97-c4ff-40ad-8dda-d5c39260b209"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1000x1000 with 16 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# visualizing the performance of the model\n",
        "sample_inputs, sample_masks, sample_truths = next(Train_dataiter)\n",
        "\n",
        "sample_inputs, sample_masks = sample_inputs.cuda(), sample_masks.cuda()\n",
        "\n",
        "model.eval()\n",
        "with torch.no_grad():\n",
        "    sample_preds = model(sample_inputs, sample_masks)\n",
        "\n",
        "fig, axs = plt.subplots(nrows=4, ncols=4, figsize=(10, 10))\n",
        "\n",
        "sample_inputs, sample_masks = sample_inputs.cpu(), sample_masks.cpu()\n",
        "for i in range(4):\n",
        "    axs[i][0].imshow(np.transpose(sample_truths[i], (1, 2, 0)))\n",
        "    axs[i][1].imshow(np.transpose(sample_masks[i], (1, 2, 0)))\n",
        "    axs[i][2].imshow(np.transpose(sample_inputs[i],(1, 2, 0)))\n",
        "    axs[i][3].imshow(np.transpose(sample_preds[i].cpu(), (1, 2, 0)))"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [
        "fbSjenKB93z9"
      ],
      "gpuType": "T4",
      "include_colab_link": true,
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}