patchify_for_inference.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Loading and embedding\n",
    "In this notebook, we do\n",
    "1. load toyzero raw window data\n",
    "1. generate masks\n",
    "1. patchify and embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.colors as mcolors\n",
    "from mpl_toolkits.axes_grid1 import make_axes_locatable\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torch.autograd import Variable\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "from transformer import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Paired Toyzero data\n",
    "Since this dataset was originally the test dataset of cycleGAN and used exclusively for test purpose, in order to use it as both train and test data for this supervised training, we set the first $1000$ pair of images for training and the remaining for test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_image_fnames(dirname):\n",
    "    \"\"\"\n",
    "    load image fnames.\n",
    "    If amax_dataset_size is not infinity and is less than all available images,\n",
    "    return a random subset of max_dataset_size image fnames.\n",
    "    \"\"\"\n",
    "    assert Path(dirname).exists(), f\"{dirname} doesn't exist\"\n",
    "    image_fnames = np.array(sorted(list(Path(dirname).glob('*npz'))))\n",
    "    return image_fnames\n",
    "\n",
    "\n",
    "class ToyzeroAlignedDataset(Dataset):\n",
    "    def __init__(\n",
    "            self, \n",
    "            dataroot, *, \n",
    "            max_dataset_size, \n",
    "            serial_batches, \n",
    "            train\n",
    "    ):\n",
    "        super(ToyzeroAlignedDataset, self).__init__()\n",
    "\n",
    "        self.serial_batches = serial_batches\n",
    "\n",
    "        dir_A = Path(dataroot)/f'testA'\n",
    "        dir_B = Path(dataroot)/f'testB'\n",
    "        self.image_fnames_A = load_image_fnames(dir_A)\n",
    "        self.image_fnames_B = load_image_fnames(dir_B)\n",
    "        \n",
    "        if train:\n",
    "            self.image_fnames_A = self.image_fnames_A[:1000]\n",
    "            self.image_fnames_B = self.image_fnames_B[:1000]\n",
    "        else:\n",
    "            self.image_fnames_A = self.image_fnames_A[1000:]\n",
    "            self.image_fnames_B = self.image_fnames_B[1000:]\n",
    "\n",
    "        assert len(self.image_fnames_A) == len(self.image_fnames_B), \"The dataset is not aligned\"\n",
    "        \n",
    "        aligned = True\n",
    "        for fname_A, fname_B in zip(self.image_fnames_A, self.image_fnames_B):\n",
    "            if fname_A.stem != fname_B.stem:\n",
    "                aligned = False\n",
    "                break\n",
    "        if not aligned:\n",
    "            print(\"The dataset is not aligned\")\n",
    "            exit()\n",
    "\n",
    "        self.size = len(self.image_fnames_A)\n",
    "\n",
    "        if max_dataset_size != float('inf') and max_dataset_size < self.size:\n",
    "            indices = np.arange(self.size)\n",
    "            np.random.shuffle(indices)\n",
    "            indices = indices[:max_dataset_size]\n",
    "            self.image_fnames_A = self.image_fnames_A[indices]\n",
    "            self.image_fnames_B = self.image_fnames_B[indices]\n",
    "            self.size = max_dataset_size\n",
    "\n",
    "    def __len__(self):\n",
    "        return self.size\n",
    "\n",
    "    def __load(self, image_fname):\n",
    "        image = np.load(image_fname)\n",
    "        image = image[image.files[0]]\n",
    "        image = np.expand_dims(np.float32(image), 0)\n",
    "        return image\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        if self.serial_batches:\n",
    "            index = index % self.size\n",
    "        else:\n",
    "            index = np.random.randint(0, self.size - 1)\n",
    "        \n",
    "        image_A = self.__load(self.image_fnames_A[index])\n",
    "        image_B = self.__load(self.image_fnames_B[index])\n",
    "\n",
    "        \n",
    "        return {\n",
    "            'A': image_A, \n",
    "            'B': image_B, \n",
    "        }\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Patchify and depatchify"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([12, 1, 5, 5])\n",
      "tensor([[[[8, 0, 3, 5, 7, 1, 0, 9, 8, 3, 8, 7, 1, 6, 3],\n",
      "          [0, 7, 0, 7, 3, 4, 0, 0, 9, 7, 8, 5, 0, 2, 8],\n",
      "          [3, 3, 5, 5, 1, 0, 1, 1, 2, 3, 5, 4, 4, 9, 4],\n",
      "          [8, 4, 9, 2, 4, 2, 2, 1, 2, 6, 6, 0, 8, 1, 9],\n",
      "          [7, 5, 1, 3, 5, 4, 1, 4, 3, 5, 0, 3, 6, 9, 0],\n",
      "          [9, 3, 8, 6, 4, 6, 8, 2, 5, 7, 3, 9, 6, 5, 6],\n",
      "          [5, 1, 4, 6, 4, 3, 3, 2, 3, 8, 7, 9, 5, 7, 6],\n",
      "          [5, 4, 2, 8, 3, 5, 1, 3, 5, 6, 1, 0, 1, 3, 6],\n",
      "          [8, 2, 9, 8, 8, 4, 4, 7, 0, 8, 5, 1, 6, 1, 5],\n",
      "          [7, 4, 0, 7, 4, 6, 1, 0, 1, 8, 4, 2, 3, 6, 7]]],\n",
      "\n",
      "\n",
      "        [[[3, 7, 3, 4, 4, 8, 2, 4, 6, 9, 3, 8, 4, 2, 3],\n",
      "          [3, 7, 2, 5, 9, 0, 0, 5, 3, 9, 0, 5, 9, 8, 2],\n",
      "          [0, 2, 1, 7, 8, 8, 8, 6, 8, 3, 4, 3, 5, 4, 8],\n",
      "          [8, 0, 9, 5, 7, 4, 8, 8, 8, 6, 8, 8, 7, 1, 8],\n",
      "          [7, 6, 5, 6, 3, 4, 5, 8, 7, 5, 8, 7, 5, 9, 9],\n",
      "          [9, 1, 1, 3, 2, 6, 1, 4, 8, 8, 3, 8, 3, 3, 4],\n",
      "          [9, 9, 4, 7, 1, 1, 4, 9, 2, 4, 4, 4, 7, 1, 8],\n",
      "          [4, 0, 2, 1, 9, 8, 4, 6, 7, 5, 9, 5, 7, 2, 1],\n",
      "          [9, 3, 7, 2, 3, 7, 1, 5, 0, 2, 0, 6, 7, 2, 0],\n",
      "          [6, 2, 5, 4, 5, 7, 1, 3, 0, 1, 7, 0, 1, 0, 8]]]])\n",
      "\n",
      "tensor([[[[8, 0, 3, 5, 7, 1, 0, 9, 8, 3, 8, 7, 1, 6, 3],\n",
      "          [0, 7, 0, 7, 3, 4, 0, 0, 9, 7, 8, 5, 0, 2, 8],\n",
      "          [3, 3, 5, 5, 1, 0, 1, 1, 2, 3, 5, 4, 4, 9, 4],\n",
      "          [8, 4, 9, 2, 4, 2, 2, 1, 2, 6, 6, 0, 8, 1, 9],\n",
      "          [7, 5, 1, 3, 5, 4, 1, 4, 3, 5, 0, 3, 6, 9, 0],\n",
      "          [9, 3, 8, 6, 4, 6, 8, 2, 5, 7, 3, 9, 6, 5, 6],\n",
      "          [5, 1, 4, 6, 4, 3, 3, 2, 3, 8, 7, 9, 5, 7, 6],\n",
      "          [5, 4, 2, 8, 3, 5, 1, 3, 5, 6, 1, 0, 1, 3, 6],\n",
      "          [8, 2, 9, 8, 8, 4, 4, 7, 0, 8, 5, 1, 6, 1, 5],\n",
      "          [7, 4, 0, 7, 4, 6, 1, 0, 1, 8, 4, 2, 3, 6, 7]]],\n",
      "\n",
      "\n",
      "        [[[3, 7, 3, 4, 4, 8, 2, 4, 6, 9, 3, 8, 4, 2, 3],\n",
      "          [3, 7, 2, 5, 9, 0, 0, 5, 3, 9, 0, 5, 9, 8, 2],\n",
      "          [0, 2, 1, 7, 8, 8, 8, 6, 8, 3, 4, 3, 5, 4, 8],\n",
      "          [8, 0, 9, 5, 7, 4, 8, 8, 8, 6, 8, 8, 7, 1, 8],\n",
      "          [7, 6, 5, 6, 3, 4, 5, 8, 7, 5, 8, 7, 5, 9, 9],\n",
      "          [9, 1, 1, 3, 2, 6, 1, 4, 8, 8, 3, 8, 3, 3, 4],\n",
      "          [9, 9, 4, 7, 1, 1, 4, 9, 2, 4, 4, 4, 7, 1, 8],\n",
      "          [4, 0, 2, 1, 9, 8, 4, 6, 7, 5, 9, 5, 7, 2, 1],\n",
      "          [9, 3, 7, 2, 3, 7, 1, 5, 0, 2, 0, 6, 7, 2, 0],\n",
      "          [6, 2, 5, 4, 5, 7, 1, 3, 0, 1, 7, 0, 1, 0, 8]]]])\n"
     ]
    }
   ],
   "source": [
    "def patchify(x, patch_size):\n",
    "    \"\"\"\n",
    "    Cut a batch of images into non-overlapping patches and stack them\n",
    "    Input:\n",
    "        - x (Tensor): The input tensor with shape (batch_size(N), n_channels(C), height(H), width(W));\n",
    "        - patch_size (int): the side length of the patch, which means each patch is of size (patch_size, patch_size);\n",
    "    Output:\n",
    "        Tensor of shape (batch_size * num_patches, n_channel, patch_size, patch_size).\n",
    "        The first 1 through num_patches belongs to the first image \n",
    "        and the num_patches + 1 through 2 * num_patches belongs to the second image, etc.\n",
    "    \"\"\"\n",
    "    batch_size = x.size(0)\n",
    "    height, width  = x.size(-2), x.size(-1)\n",
    "    assert height % patch_size == 0, f'height of image must be divisible by patch_size, but {height} % {patch_size} != 0!'\n",
    "    assert width % patch_size == 0, f'width of image must be divisible by patch_size, but {width} % {patch_size} != 0!'\n",
    "    \n",
    "    \n",
    "    patches = [\n",
    "        x[b, :, i: i + patch_size, j: j + patch_size].unsqueeze(0)\n",
    "        for b in range(batch_size)\n",
    "        for i in range(0, height, patch_size)\n",
    "        for j in range(0, width, patch_size)\n",
    "    ]\n",
    "    return torch.cat(patches, dim=0)\n",
    "\n",
    "\n",
    "def depatchify(x, batch_size, height=None, width=None):\n",
    "    \"\"\"\n",
    "    Assemble the patches back to a image.\n",
    "    NOTE: We assume that each patch is a square.\n",
    "    Input:\n",
    "        - x (Tensor): The input tensor with shape (batch_size * num_patches, n_channel, patch_size, patch_size);\n",
    "        - batch_size (int): number of FULL images in a batch;\n",
    "        - height (ing): height of the image; If given, assumed to be divisible by patch_size;\n",
    "        - width (int): width of the image; If given, assumed to be divisible by patch_size.\n",
    "        NOTE: If only one of height or width is given, infer the other one. If neither is given, assume image to be square\n",
    "    \"\"\"\n",
    "    patch_size = x.size(-1)\n",
    "    # Number of patches per image\n",
    "    assert x.size(0) % batch_size == 0, \\\n",
    "        f'the first dimension of input tensor ({x.size(0)}) should be divisible by batch_size ({batch_size})'\n",
    "    num_patches = x.size(0) // batch_size\n",
    "    \n",
    "    # Number of patches along each edge\n",
    "    if height:\n",
    "        assert height % patch_size == 0, \\\n",
    "            f'height ({height}) of image must be divisible by patch_size ({patch_size}'\n",
    "        num_patches_height = height // patch_size\n",
    "    if width:\n",
    "        assert width % patch_size == 0, \\\n",
    "            f'width ({width}) of image must be divisible by patch_size ({patch_size}'\n",
    "        num_patches_width = width // patch_size\n",
    "    if height and not width:\n",
    "        assert num_patches % num_patches_height, \\\n",
    "            f'number of patches ({num_patches}) must be divisible by the number of patches along the height ({num_patches_height})'\n",
    "        num_patches_width = num_patches // num_patches_height\n",
    "    elif width and not height:\n",
    "        assert num_patches % num_patches_width, \\\n",
    "            f'number of patches ({num_patches}) must be divisible by the number of patches along the width ({num_patches_width})'\n",
    "        num_patches_height = num_patches // num_patches_width\n",
    "    elif width and height:\n",
    "        assert num_patches == num_patches_height * num_patches_width, \\\n",
    "            f'the number of patches along the height ({num_patches_height}) times the number of patches along the width ({num_patches_width}) must equal total number of patches ({num_patches})'\n",
    "    else:\n",
    "        num_patches_edge = int(np.sqrt(num_patches))\n",
    "        assert num_patches_edge * num_patches_edge  == num_patches, \"cannot make a square image\"\n",
    "        num_patches_height = num_patches_edge\n",
    "        num_patches_width = num_patches_edge\n",
    "        \n",
    "    # Reshaping\n",
    "    patches = x.view(-1, num_patches, *x.shape[1:])\n",
    "    images = []\n",
    "    for I in patches:\n",
    "        # each I is for one image\n",
    "        rows = []\n",
    "        for i in range(num_patches_height):\n",
    "            start = i * num_patches_width\n",
    "            row = torch.cat([I[start + j] for j in range(num_patches_width)], dim=-1)\n",
    "            rows.append(row)\n",
    "        image = torch.cat(rows, dim=-2)\n",
    "        image = torch.unsqueeze(image, 0)\n",
    "        images.append(image)\n",
    "    return torch.cat(images, dim=0)\n",
    "\n",
    "\n",
    "# ================================= Test  ================================= \n",
    "patch_size = 5\n",
    "height, width= 10, 15\n",
    "batch_size = 2\n",
    "\n",
    "x = np.random.randint(0, 10, size=(batch_size, 1, height, width))\n",
    "x = torch.from_numpy(x)\n",
    "p = patchify(x, patch_size)\n",
    "print(p.shape)\n",
    "y = depatchify(p, batch_size, height, width)\n",
    "# print(x == y)\n",
    "print(x)\n",
    "print()\n",
    "print(y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Embedding and Generator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Embedding(nn.Module):\n",
    "    def __init__(self, d_model, image_height, image_width, patch_size, mode='conv', leading=True):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            - d_model (int): the embedding dimension\n",
    "            - image_height (int): image height\n",
    "            - image_width (int): image width\n",
    "            - patch_size (int): We assume square patches and patch_size is the side length of the square\n",
    "            - mode (str): choose one from ['plain', 'conv', 'fourier']\n",
    "                - plain: normalization + linear (2 * d_model) + activation + normalization + linear (TO BE IMPLEMENTED)\n",
    "                - conv: convolution\n",
    "                - fourier: 2d-fourier embedding (TO BE IMPLEMENTED)\n",
    "            - leading (bool): Whether there is a leading \"token\" indicating the start of a sequence\n",
    "        \"\"\"\n",
    "        super(Embedding, self).__init__()\n",
    "        self.d_model = d_model\n",
    "        self.image_height = image_height\n",
    "        self.image_width = image_width\n",
    "        self.patch_size = patch_size\n",
    "        self.leading = leading\n",
    "        if not leading:\n",
    "            self.source = False\n",
    "        \n",
    "        if mode == 'conv':\n",
    "            num_patches =  (image_height // patch_size) * (image_width // patch_size)\n",
    "            \n",
    "            num_layers = self.get_exp(patch_size)\n",
    "            if num_layers == -1:\n",
    "                raise ValueError(f\"patch_size ({patch_size}) is not a power of 2\")\n",
    "                \n",
    "            num_channel_base = 2 ** (num_layers - 1)\n",
    "            assert d_model % num_channel_base == 0, \\\n",
    "                f\"d_model ({d_mode}) is not a multiple of ({num_channel_base})\"\n",
    "            \n",
    "            conv_blocks = []\n",
    "            in_channels = 1\n",
    "            out_channels = d_model // num_channel_base\n",
    "            for i in range(num_layers - 1):\n",
    "                conv_blocks += [\n",
    "                    nn.Conv2d(in_channels, out_channels, kernel_size=4, stride=2, padding=1), \n",
    "                    nn.BatchNorm2d(out_channels), \n",
    "                    nn.ReLU(),\n",
    "                    # nn.LeakyReLU(negative_slope=.2)\n",
    "                ]\n",
    "                in_channels = out_channels\n",
    "                out_channels *= 2\n",
    "            conv_blocks.append(nn.Conv2d(in_channels, out_channels, kernel_size=4, stride=2, padding=1))\n",
    "            self.layers = nn.ModuleList(conv_blocks)\n",
    "        else:\n",
    "            raise NotImplementedError(f\"{mode} embedding is not implemented\")\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_exp(num):\n",
    "        exp = 0\n",
    "        while num % 2 == 0:\n",
    "            num, exp = num // 2, exp + 1\n",
    "        return -1 if num != 1 else exp       \n",
    "    \n",
    "    def forward(self, x):\n",
    "        # sanity check and get number of patches\n",
    "        assert x.size(-2) == self.image_height, f'input tensor height ({x.size(-2)}) not equal to image_height ({self.image_height})'\n",
    "        assert x.size(-1) == self.image_width, f'input tensor width ({x.size(-1)}) not equal to image_width ({self.image_width})'\n",
    "        num_patches = (x.size(-2) // self.patch_size) * (x.size(-1) // self.patch_size)\n",
    "        \n",
    "        # Patchify\n",
    "        p = patchify(x, self.patch_size)\n",
    "        \n",
    "        # Convolution\n",
    "        for i, layer in enumerate(self.layers):\n",
    "            p = layer(p)\n",
    "        \n",
    "        # Re-organize the convolution output into a form that is acceptable to transformer\n",
    "        p = p.squeeze()\n",
    "        tokenized_images = []\n",
    "        for start in range(0, p.size(0), num_patches):\n",
    "            tokenized_image = p[start: start + num_patches]\n",
    "            if self.leading:\n",
    "                # Now, the tokenized_image is of shape (seq_len, embedding_length), and\n",
    "                # we need to add one row of ones as the top row.\n",
    "                # The parameter 'pad' is in the order (left, right, top, bottom).\n",
    "                tokenized_image = F.pad(tokenized_image, pad=(0, 0, 1, 0), mode='constant', value=1)\n",
    "            tokenized_image = tokenized_image.unsqueeze(0)\n",
    "            tokenized_images.append(tokenized_image)\n",
    "        T = torch.cat(tokenized_images, dim=0)    \n",
    "        return T * np.sqrt(self.d_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Generator(nn.Module):\n",
    "    \"\"\"\n",
    "    Define standard linear + softmax generation step.\n",
    "    \"\"\"\n",
    "    def __init__(self, d_model, image_height, image_width, patch_size, mode='conv', leading=True):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            - d_model (int): the embedding dimension\n",
    "            - image_height (int): image height\n",
    "            - image_width (int): image width\n",
    "            - patch_size (int): We assume square patches and patch_size is the side length of the square\n",
    "            - mode (str): choose one from ['plain', 'conv', 'fourier']\n",
    "                - plain: normalization + linear (2 * d_model) + activation + normalization + linear (TO BE IMPLEMENTED)\n",
    "                - conv: convolution\n",
    "                - fourier: 2d-fourier embedding (TO BE IMPLEMENTED)\n",
    "            - leading (bool): Whether there is a leading \"token\" indicating the start of a sequence\n",
    "        \"\"\"\n",
    "        super(Generator, self).__init__()\n",
    "        self.d_model = d_model\n",
    "        self.image_height = image_height\n",
    "        self.image_width = image_width\n",
    "        self.patch_size = patch_size\n",
    "        self.leading = leading\n",
    "        if mode == 'conv':\n",
    "            num_patches =  (image_height // patch_size) * (image_width // patch_size)\n",
    "            \n",
    "            num_layers = self.get_exp(patch_size)\n",
    "            if num_layers == -1:\n",
    "                raise ValueError(f\"patch_size ({patch_size}) is not a power of 2\")\n",
    "                \n",
    "            num_channel_base = 2 ** (num_layers - 1)\n",
    "            assert d_model % num_channel_base == 0, \\\n",
    "                f\"d_model ({d_mode}) is not a multiple of ({num_channel_base})\"\n",
    "            \n",
    "            conv_blocks = []\n",
    "            in_channels = d_model\n",
    "            for i in range(num_layers - 1):\n",
    "                out_channels = in_channels // 2\n",
    "                conv_blocks += [\n",
    "                    nn.ConvTranspose2d(in_channels, out_channels, kernel_size=4, stride=2, padding=1), \n",
    "                    nn.BatchNorm2d(out_channels), \n",
    "                    nn.ReLU(),\n",
    "                    # nn.LeakyReLU(negative_slope=.2)\n",
    "                ]\n",
    "                in_channels = out_channels\n",
    "            conv_blocks.append(nn.ConvTranspose2d(in_channels, 1, kernel_size=4, stride=2, padding=1))\n",
    "            self.layers = nn.ModuleList(conv_blocks)\n",
    "        else:\n",
    "            raise NotImplementedError(f\"{mode} embedding is not implemented\")\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_exp(num):\n",
    "        exp = 0\n",
    "        while num % 2 == 0:\n",
    "            num, exp = num // 2, exp + 1\n",
    "        return -1 if num != 1 else exp\n",
    "    \n",
    "    def forward(self, x):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            - x (Tensor): tensor of shape (batch_size, seq_len, d_model)\n",
    "        \"\"\"\n",
    "        batch_size = x.size(0)\n",
    "        \n",
    "        # 1) Stack the embedded tokens from each sequence \n",
    "        #    and get a tensor of shape (batch_size x seq_len, d_model).\n",
    "        if self.leading:\n",
    "            x = torch.cat([tokenized_image[1:] for tokenized_image in x], dim=0)\n",
    "        else:\n",
    "            x = torch.cat([tokenized_image for tokenized_image in x], dim=0)\n",
    "        # 2) Add width=1 and height=1 to the end of the tensor and \n",
    "        #    get a tensor of shape (batch_size x seq_len, d_model, 1, 1).\n",
    "        x = x.unsqueeze(-1).unsqueeze(-1)\n",
    "        # 3) Apply deconvolution, decrease the number of freatures\n",
    "        #    while increasing the image size, so that the output\n",
    "        #    has shape (batch_size x seq_len, 1, patch_size, patch_size).\n",
    "        for i, layer in enumerate(self.layers):\n",
    "            x = layer(x)\n",
    "        # 4) Depatchify the patches back to images.\n",
    "        #    The seq_len is the number of patches, which is a square.\n",
    "        #    The number of patches along each side, P, is the square root of seq_len\n",
    "        #    The resulting tensor is of size \n",
    "        #    (batch_size, 1, patch_size * P, patch_size * P), \n",
    "        #    where patch_size * P is the side length of the original image.\n",
    "        x = depatchify(x, batch_size, self.image_height, self.image_width)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test the Embedding and Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([2, 64, 256])\n",
      "torch.Size([2, 1, 128, 128])\n"
     ]
    }
   ],
   "source": [
    "# Set the device, either 'cuda' or 'cpu'\n",
    "device = 'cuda'\n",
    "\n",
    "# Load data\n",
    "dataroot='/sdcc/u/yhuang2/PROJs/GAN/datasets/ls4gan/toyzero_cropped/toyzero_1001_300-128x128_U/'\n",
    "max_dataset_size = 2\n",
    "serial_batches = True\n",
    "# batch_size = max_dataset_size\n",
    "batch_size = 2\n",
    "\n",
    "dataset = ToyzeroAlignedDataset(\n",
    "    dataroot, \n",
    "    max_dataset_size=max_dataset_size, \n",
    "    serial_batches=serial_batches, \n",
    "    # patch_size=16\n",
    "    train=True\n",
    ")\n",
    "loader = DataLoader(dataset, batch_size=batch_size, shuffle=serial_batches)\n",
    "data = next(iter(loader))\n",
    "\n",
    "# Set parameter\n",
    "image_size = 128\n",
    "patch_size = 16\n",
    "d_model = 256\n",
    "leading = False\n",
    "\n",
    "# Embedding\n",
    "eb = Embedding(d_model, image_size, image_size, patch_size, mode='conv', leading=leading).to(device)\n",
    "embedding = eb(data['A'].to(device))\n",
    "print(embedding.shape)\n",
    "\n",
    "# Generator\n",
    "gr = Generator(d_model, image_size, image_size, patch_size, mode='conv', leading=leading).to(device)\n",
    "images = gr(embedding)\n",
    "print(images.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training the Transformer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_model(\n",
    "    image_size,\n",
    "    patch_size, *,\n",
    "    N=6,\n",
    "    d_model=512,\n",
    "    d_ff=2048,\n",
    "    h=8,\n",
    "    dropout=.1, \n",
    "    mode='conv',\n",
    "    leading=True,\n",
    "):\n",
    "    \"\"\"\n",
    "    Construct a model from hyperparameters\n",
    "    Input:\n",
    "        - image_size (int): images are supposed to be square, and the image_size is the side length.\n",
    "        - patch_size (int): patches are supposed to be square, and the patch_size is the side length.\n",
    "        - N (int): number of encoder and decoder blocks.\n",
    "        - d_model (int): d_model is the embedding length of each token. d_model must be divisible by h (number of heads).\n",
    "        - d_ff (int): hidden units in the hidden layer of the feed-forward block.\n",
    "        - h (int): number of heads in the multiheaded attention.\n",
    "        - dropout (float): dropout rate used in many places in the network.\n",
    "        - mode (string): way of embedding and generator.\n",
    "        - leading (bool): whether to add a leading token to the embedded sequence (sequence of patches in our case).\n",
    "    \"\"\"\n",
    "\n",
    "    C = copy.deepcopy\n",
    "\n",
    "    attn = MultiHeadedAttention(h, d_model)\n",
    "    ff = PositionWiseFeedForward(d_model, d_ff, dropout)\n",
    "    position = PositionalEncoding(d_model, dropout)\n",
    "\n",
    "    encoder = Encoder(EncoderLayer(d_model, C(attn), C(ff), dropout), N)\n",
    "    decoder = Decoder(DecoderLayer(d_model, C(attn), C(attn), C(ff), dropout), N)\n",
    "    src_embedding = Embedding(d_model, image_size, patch_size, mode=mode, leading=leading, source=True)\n",
    "    tgt_embedding = Embedding(d_model, image_size, patch_size, mode=mode, leading=leading, source=False)\n",
    "    model = EncoderDecoder(\n",
    "        encoder,\n",
    "        decoder,\n",
    "        nn.Sequential(src_embedding, C(position)),\n",
    "        nn.Sequential(tgt_embedding, C(position)),\n",
    "        Generator(d_model, image_size, patch_size, mode=mode)#, leading=leading)\n",
    "    )\n",
    "\n",
    "    # (author of this blog): This was important from their code.\n",
    "    # Initialize parameters with Glorot / fan_avg.\n",
    "    for p in model.parameters():\n",
    "        if p.dim() > 1:\n",
    "            nn.init.xavier_uniform_(p)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Utility functions for training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def SimplePlot(x, out, y, kwargs=None):\n",
    "    \n",
    "    num_samples = 5\n",
    "    cmap = 'bwr'\n",
    "    if kwargs:\n",
    "        if num_samples in kwargs:\n",
    "            num_samples = kwargs['num_samples']\n",
    "        if cmap in kwargs:\n",
    "            cmap = kwargs['cmap']\n",
    "    \n",
    "    total_samples = x.size(0)\n",
    "    indices = np.random.choice(total_samples, num_samples, replace=False)\n",
    "    images = torch.stack([x[indices], out[indices], y[indices]]).squeeze().transpose_(0, 1).detach().cpu().numpy()\n",
    "    \n",
    "    width = 4\n",
    "    fig, axes = plt.subplots(num_samples, 3, figsize=(3 * width * 1.1, num_samples * width))\n",
    "    for i, row in enumerate(axes):\n",
    "        for j, ax in enumerate(row):\n",
    "            image = images[i][j]\n",
    "            vmin, vmax = image.min(), image.max()\n",
    "            if vmin == 0:\n",
    "                vmin = -.05\n",
    "            if vmax == 0:\n",
    "                vmax = .05\n",
    "            divnorm = mcolors.TwoSlopeNorm(vmin=vmin, vcenter=0., vmax=vmax)\n",
    "            ax.set_aspect(1)\n",
    "            im = ax.pcolormesh(images[i][j], cmap=cmap, norm=divnorm)\n",
    "            divider = make_axes_locatable(ax)\n",
    "            cax = divider.append_axes('right', size='5%', pad=0.1)\n",
    "            fig.colorbar(im, cax=cax, orientation='vertical')\n",
    "    plt.tight_layout()\n",
    "    \n",
    "    # block=False for immediate plotting. \n",
    "    # Otherwise, the plots will show all at once at the end.\n",
    "    plt.show(block=False) \n",
    "    \n",
    "# # ========================== Test ==========================\n",
    "# x = torch.randn(10, 1, 128, 128)\n",
    "# out = torch.randn(10, 1, 128, 128)\n",
    "# y = torch.randn(10, 1, 128, 128)\n",
    "# plot(x, out, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_epoch(batches, model, loss_compute, logging_freq=50, plot_func=None, plot_kwargs=None):\n",
    "    \"\"\"\n",
    "    Standard training and logging function\n",
    "    \"\"\"    \n",
    "    start = time.time()\n",
    "    total_loss = 0 # cumulative loss\n",
    "    total_image = 0\n",
    "\n",
    "    for i, batch in enumerate(batches):\n",
    "        x = batch['A'].to(device)\n",
    "        y = batch['B'].to(device)\n",
    "        out = model(x, y) # run_epoch needs to be modified since the feed forward part also need \n",
    "        \n",
    "        loss = loss_compute(out, y)\n",
    "        num_images = batch['A'].size(0)\n",
    "        total_loss += loss * batch['A'].size(0)\n",
    "        total_image += num_images\n",
    "\n",
    "        if logging_freq > 0:\n",
    "            if i % logging_freq == logging_freq - 1:\n",
    "                elapsed = time.time() - start\n",
    "                print(f'\\tIteration: {i + 1}')\n",
    "                print(f'\\t\\tCurrent loss per image:\\t{loss / num_images:.6f}')\n",
    "                print(f'\\t\\tAvg. time per batch:\\t{elapsed / logging_freq:.6f}')\n",
    "                start = time.time()\n",
    "    if plot_func:\n",
    "        if plot_kwargs:\n",
    "            plot_func(x, out, y, plot_kwargs)\n",
    "        else:\n",
    "            plot_func(x, out, y) \n",
    "            \n",
    "    return total_loss / total_image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "class NoamOpt:\n",
    "    \"\"\"\n",
    "    Optim wrapper that implements learning rate.\n",
    "    \"\"\"\n",
    "    def __init__(self, model_size, factor, warmup, optimizer):\n",
    "        self.optimizer = optimizer\n",
    "        self._step = 0\n",
    "        self.warmup = warmup\n",
    "        self.factor = factor\n",
    "        self.model_size = model_size\n",
    "        self._rate = 0\n",
    "\n",
    "        self._a = self.factor * (self.model_size ** -.5)\n",
    "        self._b = self.warmup ** (-1.5)\n",
    "\n",
    "    def step(self):\n",
    "        \"\"\"\n",
    "        update parameters and rate\n",
    "        \"\"\"\n",
    "        self._step += 1\n",
    "        rate = self.rate()\n",
    "        for p in self.optimizer.param_groups:\n",
    "            p['lr'] = rate\n",
    "\n",
    "        self._rate = rate\n",
    "        self.optimizer.step()\n",
    "\n",
    "    def rate(self, step=None):\n",
    "        \"\"\"\n",
    "        Implement lr as defined above.\n",
    "        \"\"\"\n",
    "        if step is None:\n",
    "            step = self._step\n",
    "        return self._a * min(step ** (-.5), step * self._b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_lr(optimizer):\n",
    "    for param_group in optimizer.param_groups:\n",
    "        return param_group['lr']\n",
    "\n",
    "class LossCompute:\n",
    "    \"\"\"\n",
    "    A simple loss compute and train function.\n",
    "    \"\"\"\n",
    "    def __init__(self, criterion, opt=None):\n",
    "        self.criterion = criterion\n",
    "        self.opt = opt\n",
    "    \n",
    "    def __call__(self, x, y):\n",
    "        loss = self.criterion(x, y)\n",
    "        \n",
    "        if self.opt is not None:\n",
    "            loss.backward()\n",
    "            self.opt.step()\n",
    "            self.opt.optimizer.zero_grad()\n",
    "        \n",
    "        return loss.item()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parameters and data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set parameter\n",
    "dataroot='/sdcc/u/yhuang2/PROJs/GAN/datasets/ls4gan/toyzero_cropped/toyzero_1001_300-128x128_U/'\n",
    "# Set the device, either 'cuda' or 'cpu'\n",
    "device = 'cuda'\n",
    "image_size = 128\n",
    "patch_size = 16\n",
    "d_model = 512\n",
    "leading = False\n",
    "\n",
    "# Load train data\n",
    "max_dataset_size = 1000\n",
    "serial_batches = True\n",
    "# batch_size = max_dataset_size\n",
    "batch_size = 32\n",
    "\n",
    "dataset = ToyzeroAlignedDataset(\n",
    "    dataroot, \n",
    "    max_dataset_size=max_dataset_size, \n",
    "    serial_batches=serial_batches, \n",
    "    train=True\n",
    ")\n",
    "loader_train = DataLoader(dataset, batch_size=batch_size, shuffle=serial_batches)\n",
    "\n",
    "# Load valid data\n",
    "max_dataset_size = 1000\n",
    "serial_batches = True\n",
    "batch_size = max_dataset_size\n",
    "# batch_size = 32\n",
    "\n",
    "dataset = ToyzeroAlignedDataset(\n",
    "    dataroot, \n",
    "    max_dataset_size=max_dataset_size, \n",
    "    serial_batches=serial_batches, \n",
    "    train=False\n",
    ")\n",
    "loader_valid = DataLoader(dataset, batch_size=batch_size, shuffle=serial_batches)\n",
    "\n",
    "checkpoint_folder = 'checkpoints/simple'\n",
    "if not Path(checkpoint_folder).exists():\n",
    "    Path(checkpoint_folder).mkdir(parents=True)\n",
    "checkpoint_freq = 20\n",
    "\n",
    "plot_freq = 20\n",
    "\n",
    "checkpoint_folder = 'checkpoints/transformer_full'\n",
    "if not Path(checkpoint_folder).exists():\n",
    "    Path(checkpoint_folder).mkdir(parents=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of parameters = 10.25M\n",
      "Epoch 1 / 200\n",
      "\tlr:\t\t0.000000 -> 0.007813\n",
      "\ttime per epoch:\t10.986064 seconds\n",
      "\ttrain loss:\t1.309896\n",
      "\tvalid loss:\t1.336462\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 2 / 200\n",
      "\tlr:\t\t0.007813 -> 0.005524\n",
      "\ttime per epoch:\t10.977869 seconds\n",
      "\ttrain loss:\t1.223225\n",
      "\tvalid loss:\t1.305636\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 3 / 200\n",
      "\tlr:\t\t0.005524 -> 0.004511\n",
      "\ttime per epoch:\t10.694334 seconds\n",
      "\ttrain loss:\t1.211696\n",
      "\tvalid loss:\t1.296790\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 4 / 200\n",
      "\tlr:\t\t0.004511 -> 0.003906\n",
      "\ttime per epoch:\t10.711338 seconds\n",
      "\ttrain loss:\t1.204017\n",
      "\tvalid loss:\t1.293455\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 5 / 200\n",
      "\tlr:\t\t0.003906 -> 0.003494\n",
      "\ttime per epoch:\t10.819796 seconds\n",
      "\ttrain loss:\t1.185113\n",
      "\tvalid loss:\t1.239743\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 6 / 200\n",
      "\tlr:\t\t0.003494 -> 0.003189\n",
      "\ttime per epoch:\t10.835945 seconds\n",
      "\ttrain loss:\t1.101811\n",
      "\tvalid loss:\t1.139276\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 7 / 200\n",
      "\tlr:\t\t0.003189 -> 0.002953\n",
      "\ttime per epoch:\t10.725413 seconds\n",
      "\ttrain loss:\t1.035571\n",
      "\tvalid loss:\t1.082882\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 8 / 200\n",
      "\tlr:\t\t0.002953 -> 0.002762\n",
      "\ttime per epoch:\t10.725984 seconds\n",
      "\ttrain loss:\t0.980212\n",
      "\tvalid loss:\t1.024543\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 9 / 200\n",
      "\tlr:\t\t0.002762 -> 0.002604\n",
      "\ttime per epoch:\t10.759070 seconds\n",
      "\ttrain loss:\t0.906863\n",
      "\tvalid loss:\t0.902359\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 10 / 200\n",
      "\tlr:\t\t0.002604 -> 0.002471\n",
      "\ttime per epoch:\t10.879583 seconds\n",
      "\ttrain loss:\t0.811240\n",
      "\tvalid loss:\t0.797832\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 11 / 200\n",
      "\tlr:\t\t0.002471 -> 0.002356\n",
      "\ttime per epoch:\t10.819046 seconds\n",
      "\ttrain loss:\t0.696441\n",
      "\tvalid loss:\t0.684954\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 12 / 200\n",
      "\tlr:\t\t0.002356 -> 0.002255\n",
      "\ttime per epoch:\t10.709359 seconds\n",
      "\ttrain loss:\t0.618677\n",
      "\tvalid loss:\t0.595941\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 13 / 200\n",
      "\tlr:\t\t0.002255 -> 0.002167\n",
      "\ttime per epoch:\t10.743071 seconds\n",
      "\ttrain loss:\t0.555172\n",
      "\tvalid loss:\t0.536111\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 14 / 200\n",
      "\tlr:\t\t0.002167 -> 0.002088\n",
      "\ttime per epoch:\t10.710242 seconds\n",
      "\ttrain loss:\t0.505742\n",
      "\tvalid loss:\t0.513079\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 15 / 200\n",
      "\tlr:\t\t0.002088 -> 0.002017\n",
      "\ttime per epoch:\t10.796100 seconds\n",
      "\ttrain loss:\t0.460367\n",
      "\tvalid loss:\t0.467740\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 16 / 200\n",
      "\tlr:\t\t0.002017 -> 0.001953\n",
      "\ttime per epoch:\t10.795850 seconds\n",
      "\ttrain loss:\t0.464124\n",
      "\tvalid loss:\t0.460618\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 17 / 200\n",
      "\tlr:\t\t0.001953 -> 0.001895\n",
      "\ttime per epoch:\t10.726089 seconds\n",
      "\ttrain loss:\t0.430100\n",
      "\tvalid loss:\t0.437791\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 18 / 200\n",
      "\tlr:\t\t0.001895 -> 0.001841\n",
      "\ttime per epoch:\t10.815578 seconds\n",
      "\ttrain loss:\t0.429972\n",
      "\tvalid loss:\t0.405667\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 19 / 200\n",
      "\tlr:\t\t0.001841 -> 0.001792\n",
      "\ttime per epoch:\t10.735825 seconds\n",
      "\ttrain loss:\t0.400633\n",
      "\tvalid loss:\t0.388511\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 20 / 200\n",
      "\tlr:\t\t0.001792 -> 0.001747\n",
      "\ttime per epoch:\t10.829660 seconds\n",
      "\ttrain loss:\t0.387742\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 950.4x1440 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tvalid loss:\t0.371071\n",
      "\tSaving checkpoint\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 21 / 200\n",
      "\tlr:\t\t0.001747 -> 0.001705\n",
      "\ttime per epoch:\t10.787215 seconds\n",
      "\ttrain loss:\t0.389209\n",
      "\tvalid loss:\t0.411473\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 22 / 200\n",
      "\tlr:\t\t0.001705 -> 0.001666\n",
      "\ttime per epoch:\t10.691862 seconds\n",
      "\ttrain loss:\t0.353526\n",
      "\tvalid loss:\t0.342864\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 23 / 200\n",
      "\tlr:\t\t0.001666 -> 0.001629\n",
      "\ttime per epoch:\t10.747574 seconds\n",
      "\ttrain loss:\t0.350069\n",
      "\tvalid loss:\t0.367695\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 24 / 200\n",
      "\tlr:\t\t0.001629 -> 0.001595\n",
      "\ttime per epoch:\t10.769535 seconds\n",
      "\ttrain loss:\t0.352552\n",
      "\tvalid loss:\t0.325432\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 25 / 200\n",
      "\tlr:\t\t0.001595 -> 0.001563\n",
      "\ttime per epoch:\t10.823868 seconds\n",
      "\ttrain loss:\t0.341311\n",
      "\tvalid loss:\t0.341284\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 26 / 200\n",
      "\tlr:\t\t0.001563 -> 0.001532\n",
      "\ttime per epoch:\t10.777700 seconds\n",
      "\ttrain loss:\t0.330077\n",
      "\tvalid loss:\t0.319825\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 27 / 200\n",
      "\tlr:\t\t0.001532 -> 0.001504\n",
      "\ttime per epoch:\t10.827102 seconds\n",
      "\ttrain loss:\t0.338087\n",
      "\tvalid loss:\t0.342952\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 28 / 200\n",
      "\tlr:\t\t0.001504 -> 0.001476\n",
      "\ttime per epoch:\t10.766811 seconds\n",
      "\ttrain loss:\t0.316897\n",
      "\tvalid loss:\t0.382070\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 29 / 200\n",
      "\tlr:\t\t0.001476 -> 0.001451\n",
      "\ttime per epoch:\t10.812877 seconds\n",
      "\ttrain loss:\t0.343173\n",
      "\tvalid loss:\t0.315673\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 30 / 200\n",
      "\tlr:\t\t0.001451 -> 0.001426\n",
      "\ttime per epoch:\t10.768838 seconds\n",
      "\ttrain loss:\t0.299971\n",
      "\tvalid loss:\t0.293241\n",
      "\tpeak memory use:\t25.810G\n",
      "Epoch 31 / 200\n"
     ]
    }
   ],
   "source": [
    "# model = make_model(image_size, patch_size, N=3, d_model=256).to(device)\n",
    "# num_params = sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
    "# print(f'number of parameters = {num_params/1024**2:.2f}M')\n",
    "\n",
    "# model_opt = NoamOpt(\n",
    "#     d_model,\n",
    "#     1, # factor\n",
    "#     20, # warmup\n",
    "#     torch.optim.Adam(model.parameters(), lr=0, betas=(.9, .98), eps=1e-9)\n",
    "# )\n",
    "\n",
    "# epochs = 200\n",
    "# for epoch in range(epochs):\n",
    "#     print(f'Epoch {epoch + 1} / {epochs}')\n",
    "#     # tells your model that you are training the model. \n",
    "#     # So effectively layers like dropout, batchnorm etc. \n",
    "#     # which behave different on the train and test procedures \n",
    "#     # know what is going on and hence can behave accordingly.\n",
    "#     model.train()\n",
    "#     start = time.time()\n",
    "#     old_lr = get_lr(model_opt.optimizer)\n",
    "#     loss_train = run_epoch(\n",
    "#         loader_train,\n",
    "#         model,\n",
    "#         LossCompute(nn.L1Loss(), model_opt),\n",
    "#         logging_freq=-1\n",
    "#     )\n",
    "#     new_lr = get_lr(model_opt.optimizer)\n",
    "#     print(f'\\tlr:\\t\\t{old_lr:.6f} -> {new_lr:.6f}')\n",
    "#     print(f'\\ttime per epoch:\\t{time.time() - start:.6f} seconds')\n",
    "#     print(f'\\ttrain loss:\\t{loss_train:.6f}')\n",
    "    \n",
    "#     # Validation:\n",
    "#     if epoch % plot_freq == plot_freq - 1:\n",
    "#         plot_func = SimplePlot\n",
    "#     else:\n",
    "#         plot_func = None\n",
    "    \n",
    "#     loss_valid = run_epoch(\n",
    "#         loader_valid,\n",
    "#         model,\n",
    "#         LossCompute(nn.L1Loss()), # don't use an optimizer and the loss won't be back propagated\n",
    "#         logging_freq=-1, \n",
    "#         plot_func=plot_func,\n",
    "#         plot_kwargs={'num_samples': 3}\n",
    "#     )\n",
    "#     print(f'\\tvalid loss:\\t{loss_valid:.6f}')\n",
    "    \n",
    "#     if epoch % checkpoint_freq == checkpoint_freq - 1:\n",
    "#         print('\\tSaving checkpoint')\n",
    "#         fname = f'model_dict_{epoch + 1}'\n",
    "#         torch.save(model.state_dict(), f'{checkpoint_folder}/{fname}.pt')\n",
    "#         torch.save(model.state_dict(), f'{checkpoint_folder}/model_dict_last.pt')\n",
    "        \n",
    "#     memory_cuda = torch.cuda.max_memory_allocated(device='cuda')\n",
    "#     print(f'\\tpeak memory use:\\t{memory_cuda/1024 ** 3:.3f}G')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Greedy Decoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def greedy_decode(model, x):\n",
    "    memory = model.encode(src)\n",
    "    y = "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "yi_test",
   "language": "python",
   "name": "yi_test"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}