Created
March 31, 2023 16:21
-
-
Save pigghead/2d90da0ce83250eb0d3bfdc4b8af00a3 to your computer and use it in GitHub Desktop.
DCGAN-SimpleImplementation.ipynb
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| { | |
| "nbformat": 4, | |
| "nbformat_minor": 0, | |
| "metadata": { | |
| "colab": { | |
| "provenance": [], | |
| "authorship_tag": "ABX9TyPmQr3EPD0CZLWuDNpws92m", | |
| "include_colab_link": true | |
| }, | |
| "kernelspec": { | |
| "name": "python3", | |
| "display_name": "Python 3" | |
| }, | |
| "language_info": { | |
| "name": "python" | |
| }, | |
| "accelerator": "GPU", | |
| "widgets": { | |
| "application/vnd.jupyter.widget-state+json": { | |
| "84b2967b5937443095e0cb89c5519704": { | |
| "model_module": "@jupyter-widgets/controls", | |
| "model_name": "HBoxModel", | |
| "model_module_version": "1.5.0", | |
| "state": { | |
| "_view_name": "HBoxView", | |
| "_dom_classes": [], | |
| "_model_name": "HBoxModel", | |
| "_view_module": "@jupyter-widgets/controls", | |
| "_model_module_version": "1.5.0", | |
| "_view_count": null, | |
| "_view_module_version": "1.5.0", | |
| "box_style": "", | |
| "layout": "IPY_MODEL_78459a4f65314489831c725d83c0877a", | |
| "_model_module": "@jupyter-widgets/controls", | |
| "children": [ | |
| "IPY_MODEL_2e709235ea2b42dab063e14f08e6383f", | |
| "IPY_MODEL_67a1305205db49bcac1a3be7f177282d" | |
| ] | |
| } | |
| }, | |
| "78459a4f65314489831c725d83c0877a": { | |
| "model_module": "@jupyter-widgets/base", | |
| "model_name": "LayoutModel", | |
| "model_module_version": "1.2.0", | |
| "state": { | |
| "_view_name": "LayoutView", | |
| "grid_template_rows": null, | |
| "right": null, | |
| "justify_content": null, | |
| "_view_module": "@jupyter-widgets/base", | |
| "overflow": null, | |
| "_model_module_version": "1.2.0", | |
| "_view_count": null, | |
| "flex_flow": null, | |
| "width": null, | |
| "min_width": null, | |
| "border": null, | |
| "align_items": null, | |
| "bottom": null, | |
| "_model_module": "@jupyter-widgets/base", | |
| "top": null, | |
| "grid_column": null, | |
| "overflow_y": null, | |
| "overflow_x": null, | |
| "grid_auto_flow": null, | |
| "grid_area": null, | |
| "grid_template_columns": null, | |
| "flex": null, | |
| "_model_name": "LayoutModel", | |
| "justify_items": null, | |
| "grid_row": null, | |
| "max_height": null, | |
| "align_content": null, | |
| "visibility": null, | |
| "align_self": null, | |
| "height": null, | |
| "min_height": null, | |
| "padding": null, | |
| "grid_auto_rows": null, | |
| "grid_gap": null, | |
| "max_width": null, | |
| "order": null, | |
| "_view_module_version": "1.2.0", | |
| "grid_template_areas": null, | |
| "object_position": null, | |
| "object_fit": null, | |
| "grid_auto_columns": null, | |
| "margin": null, | |
| "display": null, | |
| "left": null | |
| } | |
| }, | |
| "2e709235ea2b42dab063e14f08e6383f": { | |
| "model_module": "@jupyter-widgets/controls", | |
| "model_name": "FloatProgressModel", | |
| "model_module_version": "1.5.0", | |
| "state": { | |
| "_view_name": "ProgressView", | |
| "style": "IPY_MODEL_8e9c9f422b964f118c65b0c28ccb4ab4", | |
| "_dom_classes": [], | |
| "description": " 0%", | |
| "_model_name": "FloatProgressModel", | |
| "bar_style": "danger", | |
| "max": 136, | |
| "_view_module": "@jupyter-widgets/controls", | |
| "_model_module_version": "1.5.0", | |
| "value": 0, | |
| "_view_count": null, | |
| "_view_module_version": "1.5.0", | |
| "orientation": "horizontal", | |
| "min": 0, | |
| "description_tooltip": null, | |
| "_model_module": "@jupyter-widgets/controls", | |
| "layout": "IPY_MODEL_33606ed261614b998d9f2c0e08b54963" | |
| } | |
| }, | |
| "67a1305205db49bcac1a3be7f177282d": { | |
| "model_module": "@jupyter-widgets/controls", | |
| "model_name": "HTMLModel", | |
| "model_module_version": "1.5.0", | |
| "state": { | |
| "_view_name": "HTMLView", | |
| "style": "IPY_MODEL_4b67b60e414f435b99169c408ab52ce5", | |
| "_dom_classes": [], | |
| "description": "", | |
| "_model_name": "HTMLModel", | |
| "placeholder": "", | |
| "_view_module": "@jupyter-widgets/controls", | |
| "_model_module_version": "1.5.0", | |
| "value": " 0/136 [00:01<?, ?it/s]", | |
| "_view_count": null, | |
| "_view_module_version": "1.5.0", | |
| "description_tooltip": null, | |
| "_model_module": "@jupyter-widgets/controls", | |
| "layout": "IPY_MODEL_b414b40e11e84e4f98d5e1b72494fdc6" | |
| } | |
| }, | |
| "8e9c9f422b964f118c65b0c28ccb4ab4": { | |
| "model_module": "@jupyter-widgets/controls", | |
| "model_name": "ProgressStyleModel", | |
| "model_module_version": "1.5.0", | |
| "state": { | |
| "_view_name": "StyleView", | |
| "_model_name": "ProgressStyleModel", | |
| "description_width": "initial", | |
| "_view_module": "@jupyter-widgets/base", | |
| "_model_module_version": "1.5.0", | |
| "_view_count": null, | |
| "_view_module_version": "1.2.0", | |
| "bar_color": null, | |
| "_model_module": "@jupyter-widgets/controls" | |
| } | |
| }, | |
| "33606ed261614b998d9f2c0e08b54963": { | |
| "model_module": "@jupyter-widgets/base", | |
| "model_name": "LayoutModel", | |
| "model_module_version": "1.2.0", | |
| "state": { | |
| "_view_name": "LayoutView", | |
| "grid_template_rows": null, | |
| "right": null, | |
| "justify_content": null, | |
| "_view_module": "@jupyter-widgets/base", | |
| "overflow": null, | |
| "_model_module_version": "1.2.0", | |
| "_view_count": null, | |
| "flex_flow": null, | |
| "width": null, | |
| "min_width": null, | |
| "border": null, | |
| "align_items": null, | |
| "bottom": null, | |
| "_model_module": "@jupyter-widgets/base", | |
| "top": null, | |
| "grid_column": null, | |
| "overflow_y": null, | |
| "overflow_x": null, | |
| "grid_auto_flow": null, | |
| "grid_area": null, | |
| "grid_template_columns": null, | |
| "flex": null, | |
| "_model_name": "LayoutModel", | |
| "justify_items": null, | |
| "grid_row": null, | |
| "max_height": null, | |
| "align_content": null, | |
| "visibility": null, | |
| "align_self": null, | |
| "height": null, | |
| "min_height": null, | |
| "padding": null, | |
| "grid_auto_rows": null, | |
| "grid_gap": null, | |
| "max_width": null, | |
| "order": null, | |
| "_view_module_version": "1.2.0", | |
| "grid_template_areas": null, | |
| "object_position": null, | |
| "object_fit": null, | |
| "grid_auto_columns": null, | |
| "margin": null, | |
| "display": null, | |
| "left": null | |
| } | |
| }, | |
| "4b67b60e414f435b99169c408ab52ce5": { | |
| "model_module": "@jupyter-widgets/controls", | |
| "model_name": "DescriptionStyleModel", | |
| "model_module_version": "1.5.0", | |
| "state": { | |
| "_view_name": "StyleView", | |
| "_model_name": "DescriptionStyleModel", | |
| "description_width": "", | |
| "_view_module": "@jupyter-widgets/base", | |
| "_model_module_version": "1.5.0", | |
| "_view_count": null, | |
| "_view_module_version": "1.2.0", | |
| "_model_module": "@jupyter-widgets/controls" | |
| } | |
| }, | |
| "b414b40e11e84e4f98d5e1b72494fdc6": { | |
| "model_module": "@jupyter-widgets/base", | |
| "model_name": "LayoutModel", | |
| "model_module_version": "1.2.0", | |
| "state": { | |
| "_view_name": "LayoutView", | |
| "grid_template_rows": null, | |
| "right": null, | |
| "justify_content": null, | |
| "_view_module": "@jupyter-widgets/base", | |
| "overflow": null, | |
| "_model_module_version": "1.2.0", | |
| "_view_count": null, | |
| "flex_flow": null, | |
| "width": null, | |
| "min_width": null, | |
| "border": null, | |
| "align_items": null, | |
| "bottom": null, | |
| "_model_module": "@jupyter-widgets/base", | |
| "top": null, | |
| "grid_column": null, | |
| "overflow_y": null, | |
| "overflow_x": null, | |
| "grid_auto_flow": null, | |
| "grid_area": null, | |
| "grid_template_columns": null, | |
| "flex": null, | |
| "_model_name": "LayoutModel", | |
| "justify_items": null, | |
| "grid_row": null, | |
| "max_height": null, | |
| "align_content": null, | |
| "visibility": null, | |
| "align_self": null, | |
| "height": null, | |
| "min_height": null, | |
| "padding": null, | |
| "grid_auto_rows": null, | |
| "grid_gap": null, | |
| "max_width": null, | |
| "order": null, | |
| "_view_module_version": "1.2.0", | |
| "grid_template_areas": null, | |
| "object_position": null, | |
| "object_fit": null, | |
| "grid_auto_columns": null, | |
| "margin": null, | |
| "display": null, | |
| "left": null | |
| } | |
| } | |
| } | |
| } | |
| }, | |
| "cells": [ | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "view-in-github", | |
| "colab_type": "text" | |
| }, | |
| "source": [ | |
| "<a href=\"https://colab.research.google.com/gist/pigghead/2d90da0ce83250eb0d3bfdc4b8af00a3/dcgan-simpleimplementation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "VNZcgLAJ-Exe", | |
| "outputId": "904fc33c-ea1d-401d-92d4-7531764080e8" | |
| }, | |
| "source": [ | |
| "from google.colab import drive\n", | |
| "drive.mount('/content/drive', force_remount=True)" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "name": "stdout", | |
| "text": [ | |
| "Mounted at /content/drive\n" | |
| ] | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "K1nHyBGMrRz7", | |
| "outputId": "3b915685-e4e7-4493-f99a-ed2809c203e0" | |
| }, | |
| "source": [ | |
| "import torch\n", | |
| "from torch import nn\n", | |
| "from tqdm.auto import tqdm\n", | |
| "from torchvision import transforms\n", | |
| "from torchvision.datasets import MNIST # training dataset\n", | |
| "from torchvision.utils import make_grid\n", | |
| "from torch.utils.data import DataLoader\n", | |
| "import matplotlib.pyplot as plt\n", | |
| "import numpy as np\n", | |
| "torch.manual_seed(0)" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "<torch._C.Generator at 0x7ff17dcfdcb0>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 44 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "95Fjqut-BfoL", | |
| "outputId": "1c1198cd-b220-48ef-a855-e9320f8b8246" | |
| }, | |
| "source": [ | |
| "### File retrieved from separate colab file \n", | |
| "### (gist: https://gist.github.com/pigghead/ece7aea9b51b799fbb17b5f1e7c1521b#file-downscale-image-test-ipynb)\n", | |
| "# bring flwr_data.npy in as an ndarray\n", | |
| "flower_images = np.load('/content/drive/MyDrive/YEAR 5/IGME 797/flwr_data_64x64.npy')\n", | |
| "\n", | |
| "def confirm_import():\n", | |
| " assert flower_images.shape == (4323, 3, 64, 64), flower_images.shape #entire ndarray\n", | |
| " assert flower_images[0].shape == (3,64,64) # single instance\n", | |
| " assert flower_images[0].dtype == 'float'\n", | |
| " print(type(flower_images[0]))\n", | |
| " print('import successful')\n", | |
| "\n", | |
| "confirm_import()\n", | |
| "### NOTE: Shape currently is: Samples, Height, Width, Channels\n", | |
| "### desired shape is Samples, Channels, Height, Width\n", | |
| "\n", | |
| "\n", | |
| "\n", | |
| "print(flower_images.astype(np.double).dtype)" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "<class 'numpy.ndarray'>\n", | |
| "import successful\n", | |
| "float64\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "pzy0utf4rJCr", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "e7c387d5-eb5b-4675-b049-ee3832391609" | |
| }, | |
| "source": [ | |
| "\"\"\"\n", | |
| "Discriminator and Generator implementation from DCGAN paper\n", | |
| "\"\"\"\n", | |
| "\n", | |
| "class Discriminator(nn.Module):\n", | |
| " def __init__(self, channels_img, features_d):\n", | |
| " super(Discriminator, self).__init__()\n", | |
| " self.disc = nn.Sequential(\n", | |
| " # input: N x channels_img x 64 x 64\n", | |
| " nn.Conv2d(\n", | |
| " channels_img, features_d, kernel_size=4, stride=2, padding=1\n", | |
| " ),\n", | |
| " nn.LeakyReLU(0.2),\n", | |
| " # _block(in_channels, out_channels, kernel_size, stride, padding)\n", | |
| " self._block(features_d, features_d * 2, 4, 2, 1),\n", | |
| " self._block(features_d * 2, features_d * 4, 4, 2, 1),\n", | |
| " self._block(features_d * 4, features_d * 8, 4, 2, 1),\n", | |
| " # After all _block img output is 4x4 (Conv2d below makes into 1x1)\n", | |
| " nn.Conv2d(features_d * 8, 1, kernel_size=4, stride=2, padding=0),\n", | |
| " nn.Sigmoid(),\n", | |
| " )\n", | |
| "\n", | |
| " def _block(self, in_channels, out_channels, kernel_size, stride, padding):\n", | |
| " return nn.Sequential(\n", | |
| " nn.Conv2d(\n", | |
| " in_channels,\n", | |
| " out_channels,\n", | |
| " kernel_size,\n", | |
| " stride,\n", | |
| " padding,\n", | |
| " bias=False,\n", | |
| " ),\n", | |
| " #nn.BatchNorm2d(out_channels),\n", | |
| " nn.LeakyReLU(0.2),\n", | |
| " )\n", | |
| "\n", | |
| " def forward(self, x):\n", | |
| " return self.disc(x)\n", | |
| "\n", | |
| "\n", | |
| "class Generator(nn.Module):\n", | |
| " def __init__(self, channels_noise, channels_img, features_g):\n", | |
| " super(Generator, self).__init__()\n", | |
| " self.net = nn.Sequential(\n", | |
| " self._block(channels_noise, features_g*16, 4, 1, 0), # 4x4\n", | |
| " self._block(features_g*16, features_g*8, 4, 2, 1), # 8x8\n", | |
| " self._block(features_g*8, features_g*4, 4, 2, 1), # 16x16\n", | |
| " self._block(features_g*4, features_g*2, 4, 2, 1), # 32x32\n", | |
| " nn.ConvTranspose2d(features_g*2, channels_img, 4, 2, 1), #64x64\n", | |
| " nn.Tanh()\n", | |
| " )\n", | |
| "\n", | |
| "\n", | |
| " def _block(self, in_channels, out_channels, kernel_size, stride, padding):\n", | |
| " return nn.Sequential(\n", | |
| " nn.ConvTranspose2d(\n", | |
| " in_channels,\n", | |
| " out_channels,\n", | |
| " kernel_size,\n", | |
| " stride,\n", | |
| " padding,\n", | |
| " bias=False,\n", | |
| " ),\n", | |
| " #nn.BatchNorm2d(out_channels),\n", | |
| " nn.ReLU(),\n", | |
| " )\n", | |
| "\n", | |
| "\n", | |
| " def forward(self, x):\n", | |
| " return self.net(x)\n", | |
| "\n", | |
| "\n", | |
| "def initialize_weights(model):\n", | |
| " # Initializes weights according to the DCGAN paper\n", | |
| " for m in model.modules():\n", | |
| " if isinstance(m, (nn.Conv2d, nn.ConvTranspose2d, nn.BatchNorm2d)):\n", | |
| " nn.init.normal_(m.weight.data, 0.0, 0.02)\n", | |
| "\n", | |
| "\n", | |
| "def test():\n", | |
| " N, in_channels, H, W = 8, 3, 64, 64\n", | |
| " noise_dim = 100\n", | |
| " gen = Generator(noise_dim, in_channels, 8)\n", | |
| " z = torch.randn((N, noise_dim, 1, 1))\n", | |
| " print('gen(z) shape: ', gen(z).shape)\n", | |
| "\n", | |
| " disc = Discriminator(in_channels, 8)\n", | |
| " x = torch.randn((N, in_channels, H, W))\n", | |
| " print('disc(x) shape: ', disc(x).shape)\n", | |
| "\n", | |
| " print('disc(gen(z)) shape: ', disc(gen(z)).shape)\n", | |
| " ####### Expected: N=8, C=1, H=64, W=64\n", | |
| " assert gen(z).shape == (N, in_channels, H, W), gen(z).shape\n", | |
| " #assert gen(z2).shape == (N, in_channels, H, W), gen(z2).shape\n", | |
| "\n", | |
| "\n", | |
| "test()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "gen(z) shape: torch.Size([8, 3, 64, 64])\n", | |
| "disc(x) shape: torch.Size([8, 1, 1, 1])\n", | |
| "disc(gen(z)) shape: torch.Size([8, 1, 1, 1])\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "A6wAPd4cZ2iH", | |
| "outputId": "428b6a17-f4f5-4f51-f82b-f3ef5fc6f099" | |
| }, | |
| "source": [ | |
| "class discriminator_32x32(nn.Module):\n", | |
| " \n", | |
| " def __init__(self, channels_img, features_d):\n", | |
| " super(discriminator_32x32, self).__init__()\n", | |
| " self.disc = nn.Sequential(\n", | |
| " # Input: N x channels_img x features_d x features_d (?)\n", | |
| " self._block(channels_img, features_d*2, 4, 2, 1), # 16x16\n", | |
| " self._block(features_d*2, features_d*4, 4, 2, 1), #8x8\n", | |
| " self._block(features_d*4, features_d*8, 4, 2, 1), #4x4\n", | |
| " #self._block(features_d*8, features_d*16, 4, 2, 1),\n", | |
| " nn.Conv2d(features_d*8, 1, 4, 1, 0), #2x2\n", | |
| " nn.Sigmoid() # 1 or 0\n", | |
| " )\n", | |
| "\n", | |
| " def _block(self, in_channels, out_channels, kernel_size, stride, padding):\n", | |
| " return nn.Sequential(\n", | |
| " nn.Conv2d(in_channels, out_channels, kernel_size, stride, padding, bias=False),\n", | |
| " nn.BatchNorm2d(out_channels),\n", | |
| " nn.LeakyReLU(0.2, inplace=True)\n", | |
| " )\n", | |
| "\n", | |
| "\n", | |
| " def forward(self, img):\n", | |
| " return self.disc(img)\n", | |
| "\n", | |
| "''' EXAMPLE TEST CLASS\n", | |
| "\n", | |
| " N, in_channels, H, W = 8, 3, 64, 64\n", | |
| " noise_dim = 100\n", | |
| " x = torch.randn((N, in_channels, H, W))\n", | |
| "\n", | |
| " disc = Discriminator(in_channels, 8)\n", | |
| " assert disc(x).shape == (N, 1, 1, 1), \"Discriminator test failed\"\n", | |
| "'''\n", | |
| "\n", | |
| "def test_disc():\n", | |
| " N, in_channels, H, W = 8, 1, 32, 32\n", | |
| " noise_dim = 100\n", | |
| " x = torch.randn((N, in_channels, H, W)) # !!ERROR HERE (resolved)\n", | |
| " print(x.shape)\n", | |
| " disc = discriminator_32x32(in_channels, 32) \n", | |
| "\n", | |
| " print( disc(x)[0] )\n", | |
| " assert disc(x).shape == (N, 1, 1, 1), \"Discriminator test failed\"\n", | |
| "\n", | |
| "\n", | |
| "test_disc()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "torch.Size([8, 1, 32, 32])\n", | |
| "tensor([[[0.2947]]], grad_fn=<SelectBackward>)\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 283 | |
| }, | |
| "id": "pALZEKeiQVUW", | |
| "outputId": "804ba980-47a4-430a-8799-353648ba4ce9" | |
| }, | |
| "source": [ | |
| "class generator_32x32(nn.Module):\n", | |
| " \n", | |
| " def __init__(self, noise_channels, im_channels, features_g):\n", | |
| " super(generator_32x32, self).__init__()\n", | |
| " self.gen = nn.Sequential( \n", | |
| " self._block(noise_channels, features_g*8, 4, 1, 0), #4x4\n", | |
| " self._block(features_g*8, features_g*4, 4, 2, 1), #8x8\n", | |
| " self._block(features_g*4, features_g*2, 4, 2, 1), #16x16\n", | |
| " nn.ConvTranspose2d(\n", | |
| " features_g*2, im_channels, 4, 2, 1\n", | |
| " ),\n", | |
| " # Output= s(n-1)+f-2p, 32x32 img\n", | |
| " nn.Tanh()\n", | |
| " )\n", | |
| "\n", | |
| "\n", | |
| " def _block(self, in_channel, out_channel, kernel_size, stride, padding):\n", | |
| " ''' \n", | |
| " Performs transposed convolution, batch normalization, and Leaky ReLU activation \n", | |
| " '''\n", | |
| " return nn.Sequential (\n", | |
| " nn.ConvTranspose2d(in_channel, out_channel, kernel_size, stride, padding, bias=False),\n", | |
| " # Batch norm?\n", | |
| " nn.BatchNorm2d(out_channel),\n", | |
| " nn.LeakyReLU(0.2)\n", | |
| " )\n", | |
| "\n", | |
| "\n", | |
| " def forward(self, x):\n", | |
| " return self.gen(x)\n", | |
| "\n", | |
| "\n", | |
| "''' EXAMPLE TEST CLASS\n", | |
| "\n", | |
| " N, in_channels, H, W = 8, 3, 64, 64\n", | |
| " noise_dim = 100\n", | |
| " x = torch.randn((N, in_channels, H, W))\n", | |
| "\n", | |
| " disc = Discriminator(in_channels, 8)\n", | |
| " assert disc(x).shape == (N, 1, 1, 1), \"Discriminator test failed\"\n", | |
| "\n", | |
| " gen = Generator(noise_dim, in_channels, 8)\n", | |
| " z = torch.randn((N, noise_dim, 1, 1))\n", | |
| " assert gen(z).shape == (N, in_channels, H, W), \"Generator test failed\"\n", | |
| "'''\n", | |
| "\n", | |
| "\n", | |
| "# test our 28x28 generator\n", | |
| "def test_gen_28x28():\n", | |
| " N, in_channels, H, W = 8, 1, 32, 32\n", | |
| " noise_dim = 100\n", | |
| " z = torch.randn((N, noise_dim, 1, 1))\n", | |
| " gen1 = generator_32x32(noise_dim, in_channels, 32)\n", | |
| " assert gen1(z).shape == (N, in_channels, H, W)\n", | |
| " print('gen1(z).shape',gen1(z).shape)\n", | |
| "\n", | |
| "\n", | |
| "#test_gen_28x28()\n", | |
| "\n", | |
| "\n", | |
| "def print_gen_28x28():\n", | |
| " N, in_channels, H, W = 1, 1, 32, 32\n", | |
| " noise_dim = 128\n", | |
| " z = torch.randn((N, noise_dim, 1, 1),device='cpu')\n", | |
| " gen = generator_32x32(noise_dim, in_channels, 8).to('cpu')\n", | |
| " print('gen(z) shape: ', gen(z).shape)\n", | |
| " plt.Figure(figsize=(28, 28))\n", | |
| " plt.imshow(gen(z)[0, 0, :, :].detach().cpu(), cmap='gray')\n", | |
| " plt.show()\n", | |
| "\n", | |
| " \n", | |
| "#test_gen_28x28()\n", | |
| "print_gen_28x28()\n", | |
| "\n", | |
| "\n", | |
| "# test both\n", | |
| "def test_both():\n", | |
| " N, in_channels, H, W = 8, 1, 32, 32\n", | |
| " noise_dim = 100\n", | |
| " # A test generator\n", | |
| " z = torch.randn((N, noise_dim, 1, 1))\n", | |
| " gen = generator_32x32(noise_dim, in_channels, 32)\n", | |
| " # A test discriminator\n", | |
| " disc = discriminator_32x32(in_channels, 32)\n", | |
| " print(disc(gen(z)).shape)\n", | |
| "\n", | |
| "\n", | |
| "#test_both()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "gen(z) shape: torch.Size([1, 1, 32, 32])\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "display_data", | |
| "data": { | |
| "image/png": "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\n", | |
| "text/plain": [ | |
| "<Figure size 432x288 with 1 Axes>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [], | |
| "needs_background": "light" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "8-KR-tcPyqhc" | |
| }, | |
| "source": [ | |
| "def show_tensor_image(img_tensor, plotTitle, num_images=25, size=(1,32,32)):\n", | |
| " ''' Make a grid of images ''' \n", | |
| " image_tensor = (img_tensor+1)/2\n", | |
| " image_unflat = image_tensor.detach().cpu()\n", | |
| "\n", | |
| " image_grid = make_grid(image_unflat[:num_images], nrow=5)\n", | |
| "\n", | |
| " plt.title(plotTitle)\n", | |
| " plt.imshow(image_grid.permute(1, 2, 0).squeeze())\n", | |
| " plt.show()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 441, | |
| "referenced_widgets": [ | |
| "84b2967b5937443095e0cb89c5519704", | |
| "78459a4f65314489831c725d83c0877a", | |
| "2e709235ea2b42dab063e14f08e6383f", | |
| "67a1305205db49bcac1a3be7f177282d", | |
| "8e9c9f422b964f118c65b0c28ccb4ab4", | |
| "33606ed261614b998d9f2c0e08b54963", | |
| "4b67b60e414f435b99169c408ab52ce5", | |
| "b414b40e11e84e4f98d5e1b72494fdc6" | |
| ] | |
| }, | |
| "id": "aMgfXY1aqaR_", | |
| "outputId": "0162ac23-0683-41ce-8496-30d21949b5d8" | |
| }, | |
| "source": [ | |
| "### TRAINING ###\n", | |
| "criterion = nn.BCELoss()\n", | |
| "n_epochs = 500\n", | |
| "display_step = 500 # for visualization\n", | |
| "cur_step = 0\n", | |
| "noise_dim = 100\n", | |
| "IMG_SIZE = 32\n", | |
| "BATCH_SIZE = 32\n", | |
| "LEARNING_RATE = 0.0002 # a /alpha\n", | |
| "FEATURES = 32 # this must correspond to the architecture of both d and g (32x32)\n", | |
| "NUM_CHANNELS = 3\n", | |
| "device = 'cpu'\n", | |
| "\n", | |
| "# betas as described in DCGAN paper\n", | |
| "beta_1 = 0.5\n", | |
| "beta_2 = 0.9\n", | |
| "\n", | |
| "transform = transforms.Compose([\n", | |
| " transforms.Resize(IMG_SIZE),\n", | |
| " transforms.ToTensor(),\n", | |
| " transforms.Normalize(\n", | |
| " [0.5 for _ in range(1)] , \n", | |
| " [0.5 for _ in range(1)]\n", | |
| " )\n", | |
| "])\n", | |
| "\n", | |
| "d_loader = DataLoader(\n", | |
| " flower_images,\n", | |
| " batch_size = BATCH_SIZE,\n", | |
| " shuffle = True\n", | |
| ")\n", | |
| "\n", | |
| "dataloader = DataLoader(\n", | |
| " MNIST('.', train=True, download=False, transform=transform),\n", | |
| " batch_size = BATCH_SIZE,\n", | |
| " shuffle = True\n", | |
| ")\n", | |
| "\n", | |
| "# Initialize generator and discriminator (32x32 variant)\n", | |
| "gen = Generator(noise_dim, NUM_CHANNELS, 64).to(device)\n", | |
| "disc = Discriminator(NUM_CHANNELS, 64).to(device)\n", | |
| "\n", | |
| "# adam optimizers, dcgan paper\n", | |
| "gen_optim = torch.optim.Adam(gen.parameters(), lr=LEARNING_RATE, betas=(beta_1, beta_2))\n", | |
| "disc_optim = torch.optim.Adam(disc.parameters(), lr=LEARNING_RATE, betas=(beta_1, beta_2))\n", | |
| "\n", | |
| "# initialize weights\n", | |
| "#gen = gen.apply(initialize_weights)\n", | |
| "#disc = disc.apply(initialize_weights)\n", | |
| "initialize_weights(gen)\n", | |
| "initialize_weights(disc)\n", | |
| "\n", | |
| "for epochs in range(n_epochs):\n", | |
| " # we aren't using labels as of right now (StyleGAN will use these)\n", | |
| " for real in tqdm(d_loader):\n", | |
| " cur_batch_size = len(real)\n", | |
| " real = real.double().to(device)\n", | |
| " fake_noise = torch.randn((cur_batch_size, noise_dim, 1, 1), device=device)\n", | |
| " fake = gen(fake_noise)\n", | |
| "\n", | |
| " print(real.dtype)\n", | |
| "\n", | |
| " ### Updating disc\n", | |
| " disc_real_pred = disc(real)\n", | |
| " disc_real_loss = criterion(disc_real_pred, torch.ones_like(disc_real_pred))\n", | |
| "\n", | |
| " disc_fake_pred = disc(fake.detach()).reshape(-1)\n", | |
| " disc_fake_loss = criterion(disc_fake_pred, torch.zeros_like(disc_fake_pred))\n", | |
| "\n", | |
| " # total loss is the avg of the two\n", | |
| " #disc_loss = disc_real_loss - disc_fake_loss\n", | |
| " disc_loss = (disc_fake_loss + disc_real_loss) / 2\n", | |
| "\n", | |
| " # update gradients\n", | |
| " disc.zero_grad()\n", | |
| " disc_loss.backward(retain_graph=True)\n", | |
| " disc_optim.step()\n", | |
| "\n", | |
| " ### Updating gen\n", | |
| " output = disc(fake).reshape(-1)\n", | |
| " gen_loss = criterion(output, torch.ones_like(output))\n", | |
| " gen.zero_grad()\n", | |
| " gen_loss.backward()\n", | |
| " gen_optim.step()\n", | |
| "\n", | |
| " if cur_step % display_step == 0 and cur_step > 0:\n", | |
| " print(f'Step {cur_step}: Generator Loss: {gen_loss}, discriminator loss: {disc_loss}')\n", | |
| " show_tensor_image(fake, 'fake/generated', num_images=25)\n", | |
| " show_tensor_image(real, 'real/dataset', num_images=25)\n", | |
| " mean_gen_loss = 0\n", | |
| " mean_disc_loss = 0\n", | |
| " cur_step+=1\n" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "display_data", | |
| "data": { | |
| "application/vnd.jupyter.widget-view+json": { | |
| "model_id": "84b2967b5937443095e0cb89c5519704", | |
| "version_minor": 0, | |
| "version_major": 2 | |
| }, | |
| "text/plain": [ | |
| "HBox(children=(FloatProgress(value=0.0, max=136.0), HTML(value='')))" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| } | |
| }, | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "torch.float64\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "error", | |
| "ename": "RuntimeError", | |
| "evalue": "ignored", | |
| "traceback": [ | |
| "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", | |
| "\u001b[0;31mRuntimeError\u001b[0m Traceback (most recent call last)", | |
| "\u001b[0;32m<ipython-input-83-1c31cc3762f1>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m 62\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 63\u001b[0m \u001b[0;31m### Updating disc\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 64\u001b[0;31m \u001b[0mdisc_real_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdisc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreal\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 65\u001b[0m \u001b[0mdisc_real_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdisc_real_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mones_like\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdisc_real_pred\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 66\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m 887\u001b[0m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 888\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 890\u001b[0m for hook in itertools.chain(\n\u001b[1;32m 891\u001b[0m \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m<ipython-input-46-6554cda8df86>\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m 36\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 37\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 38\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdisc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 39\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 40\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m 887\u001b[0m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 888\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 890\u001b[0m for hook in itertools.chain(\n\u001b[1;32m 891\u001b[0m \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/container.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m 117\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 118\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 119\u001b[0;31m \u001b[0minput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 120\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 121\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m 887\u001b[0m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 888\u001b[0m \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 890\u001b[0m for hook in itertools.chain(\n\u001b[1;32m 891\u001b[0m \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m 397\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 398\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 399\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_conv_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbias\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 400\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 401\u001b[0m \u001b[0;32mclass\u001b[0m \u001b[0mConv3d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_ConvNd\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;32m/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py\u001b[0m in \u001b[0;36m_conv_forward\u001b[0;34m(self, input, weight, bias)\u001b[0m\n\u001b[1;32m 394\u001b[0m _pair(0), self.dilation, self.groups)\n\u001b[1;32m 395\u001b[0m return F.conv2d(input, weight, bias, self.stride,\n\u001b[0;32m--> 396\u001b[0;31m self.padding, self.dilation, self.groups)\n\u001b[0m\u001b[1;32m 397\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 398\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n", | |
| "\u001b[0;31mRuntimeError\u001b[0m: expected scalar type Double but found Float" | |
| ] | |
| }, | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "rKUG2IlWKm-q" | |
| }, | |
| "source": [], | |
| "execution_count": null, | |
| "outputs": [] | |
| } | |
| ] | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment