Resnet.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Downloading datasets and installing packages\n",
    "Uncomment to run; only necessary once for each user.\n",
    "source: https://github.com/Noob-can-Compile/Automatic-Image-Captioning/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #Gets validation data.\n",
    "#\n",
    "#!wget http://images.cocodataset.org/zips/val2014.zip\n",
    "#!unzip val2014.zip -d images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #Gets training data.\n",
    "\n",
    "# !wget  http://images.cocodataset.org/zips/train2014.zip\n",
    "# !unzip train2014.zip -d images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #downloads and unzips train/val annotations.\n",
    "\n",
    "# !wget http://images.cocodataset.org/annotations/annotations_trainval2014.zip\n",
    "# !unzip annotations_trainval2014.zip "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !wget http://images.cocodataset.org/annotations/image_info_test2014.zip\n",
    "# !unzip image_info_test2014.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #downloads and unzips test data\n",
    "# !wget http://images.cocodataset.org/zips/test2014.zip\n",
    "# !unzip 'test2014.zip' -d images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #Install missing packages if modules not loaded\n",
    "# !pip install --user torch==1.7.0+cu110 torchvision==0.8.1+cu110 torchaudio===0.7.0 -f https://download.pytorch.org/whl/torch_stable.html\n",
    "# !pip install nltk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "B68L-vRriH1y"
   },
   "outputs": [],
   "source": [
    "# #Only use in Colab\n",
    "# from google.colab import files\n",
    "\n",
    "#!cd models\n",
    "\n",
    "#files.download('encoder-1.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Master import list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "import nltk\n",
    "import numpy as np\n",
    "import pickle\n",
    "import os\n",
    "import os.path\n",
    "import random\n",
    "import requests\n",
    "import sys\n",
    "import time\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchvision.models as models\n",
    "import torch.utils.data as data\n",
    "from torchvision import transforms\n",
    "from pycocotools.coco import COCO\n",
    "from collections import Counter\n",
    "from PIL import Image\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Global definitions\n",
    "Note that filenames will depend on the model and must be changed if you don't want to overwrite an existing log file or encoder/decoder .pkl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "#nltk.download('punkt')\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "#dataloader, encoder, and decoder settings\n",
    "batch_size = 64          # batch size\n",
    "vocab_threshold = 5        # minimum word count threshold\n",
    "vocab_from_file = True    # if True, load existing vocab file\n",
    "embed_size = 300           # dimensionality of image and word embeddings\n",
    "hidden_size = 512          # number of features in hidden state of the RNN decoder\n",
    "\n",
    "#training loop\n",
    "num_epochs = 3             # number of training epochs\n",
    "save_every = 1             # determines frequency of saving model weights\n",
    "print_every = 100          # determines window for printing average loss\n",
    "\n",
    "#BE SURE TO CHANGE THESE FILENAMES IF YOU CHANGE THE CNN\n",
    "log_file = 'training_log_model0resnet50.txt'       # name of file with saved training loss and perplexity\n",
    "loss_file = \"res34losses.txt\"\n",
    "perplexity_file = \"res34perplexities.txt\"\n",
    "\n",
    "#Specify the saved models to load.\n",
    "encoder_file = 'res34encoder-3.pkl'\n",
    "decoder_file = 'res34decoder-3.pkl'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Encoder/Decoder definitions\n",
    "This model uses resnet50 as the encoder and an RNN with LSTM as the decoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "id": "Qn81ycM3c4_E"
   },
   "outputs": [],
   "source": [
    "class EncoderCNN(nn.Module):\n",
    "    def __init__(self, embed_size):\n",
    "        super(EncoderCNN, self).__init__()\n",
    "        resnet = models.resnet34(pretrained=True)\n",
    "        for param in resnet.parameters():\n",
    "            param.requires_grad_(False)\n",
    "        \n",
    "        modules = list(resnet.children())[:-1]\n",
    "        self.resnet = nn.Sequential(*modules)\n",
    "        self.embed = nn.Linear(resnet.fc.in_features, embed_size)\n",
    "\n",
    "    def forward(self, images):\n",
    "#         print(\"resnet: \", self.resnet)\n",
    "        features = self.resnet(images)\n",
    "#        print(\"out of resnet: \", features.shape)\n",
    "        features = features.view(features.size(0), -1)\n",
    "#         print(\"out of view: \", features.shape)\n",
    "        features = self.embed(features)\n",
    "#         print(\"forward output shape: \", features.shape)\n",
    "        return features\n",
    "    \n",
    "\n",
    "class DecoderRNN(nn.Module):\n",
    "    def __init__(self, embed_size, hidden_size, vocab_size, num_layers=1):\n",
    "        super().__init__()\n",
    "        self.embedding_layer = nn.Embedding(vocab_size, embed_size)\n",
    "        \n",
    "        self.lstm = nn.LSTM(input_size = embed_size,hidden_size = hidden_size,\n",
    "                            num_layers = num_layers, batch_first = True)\n",
    "        \n",
    "        self.linear = nn.Linear(hidden_size, vocab_size)\n",
    "    \n",
    "    def forward(self, features, captions):\n",
    "        captions = captions[:, :-1]\n",
    "        embed = self.embedding_layer(captions)\n",
    "        embed = torch.cat((features.unsqueeze(1), embed), dim = 1)\n",
    "        lstm_outputs, _ = self.lstm(embed)\n",
    "        out = self.linear(lstm_outputs)\n",
    "        \n",
    "        return out\n",
    "\n",
    "    def sample(self, inputs, states=None, max_len=20):\n",
    "        \" accepts pre-processed image tensor (inputs) and returns predicted sentence (list of tensor ids of length max_len) \"\n",
    "        output_sentence = []\n",
    "        for i in range(max_len):\n",
    "            lstm_outputs, states = self.lstm(inputs, states)\n",
    "            lstm_outputs = lstm_outputs.squeeze(1)\n",
    "            out = self.linear(lstm_outputs)\n",
    "            last_pick = out.max(1)[1]\n",
    "            #print(\"last_pick: \", last_pick.shape)\n",
    "            output_sentence.append(last_pick.item())\n",
    "            inputs = self.embedding_layer(last_pick).unsqueeze(1)\n",
    "        \n",
    "        return output_sentence"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VYvddgbxzMga"
   },
   "source": [
    "# Dataset and Vocabulary classes\n",
    "Uses COCO dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "id": "-h62qULznoHH"
   },
   "outputs": [],
   "source": [
    "class CoCoDataset(data.Dataset):\n",
    "    \n",
    "    def __init__(self, transform, mode, batch_size, vocab_threshold, vocab_file, start_word, \n",
    "        end_word, unk_word, annotations_file, vocab_from_file, img_folder):\n",
    "        self.transform = transform\n",
    "        self.mode = mode\n",
    "        self.batch_size = batch_size\n",
    "        self.vocab = Vocabulary(vocab_threshold, vocab_file, start_word,\n",
    "            end_word, unk_word, annotations_file, vocab_from_file)\n",
    "        self.img_folder = img_folder\n",
    "        if self.mode == 'train' or self.mode == 'val':\n",
    "            self.coco = COCO(annotations_file)\n",
    "            self.ids = list(self.coco.anns.keys())\n",
    "            print('Obtaining caption lengths...')\n",
    "            all_tokens = [nltk.tokenize.word_tokenize(str(self.coco.anns[self.ids[index]]['caption']).lower()) for index in tqdm(np.arange(len(self.ids)))]\n",
    "            self.caption_lengths = [len(token) for token in all_tokens]\n",
    "        else:\n",
    "            test_info = json.loads(open(annotations_file).read())\n",
    "            self.paths = [item['file_name'] for item in test_info['images']]\n",
    "        \n",
    "    def __getitem__(self, index):\n",
    "        # obtain image and caption if in training mode\n",
    "        if self.mode == 'train' or self.mode == 'val':\n",
    "            ann_id = self.ids[index]\n",
    "            caption = self.coco.anns[ann_id]['caption']\n",
    "            img_id = self.coco.anns[ann_id]['image_id']\n",
    "            path = self.coco.loadImgs(img_id)[0]['file_name']\n",
    "\n",
    "            # Convert image to tensor and pre-process using transform\n",
    "            image = Image.open(os.path.join(self.img_folder, path)).convert('RGB')\n",
    "            image = self.transform(image)\n",
    "\n",
    "            # Convert caption to tensor of word ids.\n",
    "            tokens = nltk.tokenize.word_tokenize(str(caption).lower())\n",
    "            caption = []\n",
    "            caption.append(self.vocab(self.vocab.start_word))\n",
    "            caption.extend([self.vocab(token) for token in tokens])\n",
    "            caption.append(self.vocab(self.vocab.end_word))\n",
    "            caption = torch.Tensor(caption).long()\n",
    "\n",
    "            # return pre-processed image and caption tensors\n",
    "            return image, caption\n",
    "\n",
    "        # obtain image if in test mode\n",
    "        else:\n",
    "            path = self.paths[index]\n",
    "\n",
    "            # Convert image to tensor and pre-process using transform\n",
    "            PIL_image = Image.open(os.path.join(self.img_folder, path)).convert('RGB')\n",
    "            orig_image = np.array(PIL_image)\n",
    "            image = self.transform(PIL_image)\n",
    "\n",
    "            # return original image and pre-processed image tensor\n",
    "            return orig_image, image\n",
    "\n",
    "    def get_train_indices(self):\n",
    "        sel_length = np.random.choice(self.caption_lengths)\n",
    "        all_indices = np.where([self.caption_lengths[i] == sel_length for i in np.arange(len(self.caption_lengths))])[0]\n",
    "        indices = list(np.random.choice(all_indices, size=self.batch_size))\n",
    "        return indices\n",
    "\n",
    "    def __len__(self):\n",
    "        if self.mode == 'train' or self.mode == 'val':\n",
    "            return len(self.ids)\n",
    "        else:\n",
    "            return len(self.paths)\n",
    "\n",
    "class Vocabulary(object):\n",
    "\n",
    "    def __init__(self,\n",
    "        vocab_threshold,\n",
    "        vocab_file='./vocab.pkl',\n",
    "        start_word=\"<start>\",\n",
    "        end_word=\"<end>\",\n",
    "        unk_word=\"<unk>\",\n",
    "        annotations_file='annotations/captions_train2014.json',\n",
    "        vocab_from_file=False):\n",
    "        \"\"\"Initialize the vocabulary.\n",
    "        Args:\n",
    "          vocab_threshold: Minimum word count threshold.\n",
    "          vocab_file: File containing the vocabulary.\n",
    "          start_word: Special word denoting sentence start.\n",
    "          end_word: Special word denoting sentence end.\n",
    "          unk_word: Special word denoting unknown words.\n",
    "          annotations_file: Path for train annotation file.\n",
    "          vocab_from_file: If False, create vocab from scratch & override any existing vocab_file\n",
    "                           If True, load vocab from from existing vocab_file, if it exists\n",
    "        \"\"\"\n",
    "        self.vocab_threshold = vocab_threshold\n",
    "        self.vocab_file = vocab_file\n",
    "        self.start_word = start_word\n",
    "        self.end_word = end_word\n",
    "        self.unk_word = unk_word\n",
    "        self.annotations_file = annotations_file\n",
    "        self.vocab_from_file = vocab_from_file\n",
    "        self.get_vocab()\n",
    "\n",
    "    def get_vocab(self):\n",
    "        \"\"\"Load the vocabulary from file OR build the vocabulary from scratch.\"\"\"\n",
    "        if os.path.exists(self.vocab_file) & self.vocab_from_file:\n",
    "            with open(self.vocab_file, 'rb') as f:\n",
    "                vocab = pickle.load(f)\n",
    "                self.word2idx = vocab.word2idx\n",
    "                self.idx2word = vocab.idx2word\n",
    "            print('Vocabulary successfully loaded from vocab.pkl file!')\n",
    "        else:\n",
    "            self.build_vocab()\n",
    "            with open(self.vocab_file, 'wb') as f:\n",
    "                pickle.dump(self, f)\n",
    "        \n",
    "    def build_vocab(self):\n",
    "        \"\"\"Populate the dictionaries for converting tokens to integers (and vice-versa).\"\"\"\n",
    "        self.init_vocab()\n",
    "        self.add_word(self.start_word)\n",
    "        self.add_word(self.end_word)\n",
    "        self.add_word(self.unk_word)\n",
    "        self.add_captions()\n",
    "\n",
    "    def init_vocab(self):\n",
    "        \"\"\"Initialize the dictionaries for converting tokens to integers (and vice-versa).\"\"\"\n",
    "        self.word2idx = {}\n",
    "        self.idx2word = {}\n",
    "        self.idx = 0\n",
    "\n",
    "    def add_word(self, word):\n",
    "        \"\"\"Add a token to the vocabulary.\"\"\"\n",
    "        if not word in self.word2idx:\n",
    "            self.word2idx[word] = self.idx\n",
    "            self.idx2word[self.idx] = word\n",
    "            self.idx += 1\n",
    "\n",
    "    def add_captions(self):\n",
    "        \"\"\"Loop over training captions and add all tokens to the vocabulary that meet or exceed the threshold.\"\"\"\n",
    "        coco = COCO(self.annotations_file)\n",
    "        counter = Counter()\n",
    "        ids = coco.anns.keys()\n",
    "        for i, id in enumerate(ids):\n",
    "            caption = str(coco.anns[id]['caption'])\n",
    "            tokens = nltk.tokenize.word_tokenize(caption.lower())\n",
    "            counter.update(tokens)\n",
    "\n",
    "            if i % 100000 == 0:\n",
    "                print(\"[%d/%d] Tokenizing captions...\" % (i, len(ids)))\n",
    "\n",
    "        words = [word for word, cnt in counter.items() if cnt >= self.vocab_threshold]\n",
    "\n",
    "        for i, word in enumerate(words):\n",
    "            self.add_word(word)\n",
    "\n",
    "    def __call__(self, word):\n",
    "        if not word in self.word2idx:\n",
    "            return self.word2idx[self.unk_word]\n",
    "        return self.word2idx[word]\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.word2idx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kZn24O1XzYKK"
   },
   "source": [
    "# Define dataloader modes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_loader(transform,\n",
    "               mode='train',\n",
    "               batch_size=1,\n",
    "               vocab_threshold=None,\n",
    "               vocab_file='./vocab.pkl',\n",
    "               start_word=\"<start>\",\n",
    "               end_word=\"<end>\",\n",
    "               unk_word=\"<unk>\",\n",
    "               vocab_from_file=True,\n",
    "               num_workers=0,\n",
    "               cocoapi_loc=''):\n",
    "    \"\"\"Returns the data loader.\n",
    "    Args:\n",
    "      transform: Image transform.\n",
    "      mode: One of 'train' or 'test'.\n",
    "      batch_size: Batch size (if in testing mode, must have batch_size=1).\n",
    "      vocab_threshold: Minimum word count threshold.\n",
    "      vocab_file: File containing the vocabulary. \n",
    "      start_word: Special word denoting sentence start.\n",
    "      end_word: Special word denoting sentence end.\n",
    "      unk_word: Special word denoting unknown words.\n",
    "      vocab_from_file: If False, create vocab from scratch & override any existing vocab_file.\n",
    "                       If True, load vocab from from existing vocab_file, if it exists.\n",
    "      num_workers: Number of subprocesses to use for data loading \n",
    "      cocoapi_loc: The location of the folder containing the COCO API: https://github.com/cocodataset/cocoapi\n",
    "    \"\"\"\n",
    "    \n",
    "    assert mode in ['train', 'val', 'test'], \"mode must be one of 'train' or 'test'.\"\n",
    "    #if vocab_from_file==False: assert mode=='train', \"To generate vocab from captions file, must be in training mode (mode='train').\"\n",
    "\n",
    "    # Based on mode (train, val, test), obtain img_folder and annotations_file.\n",
    "    if mode == 'train':\n",
    "        if vocab_from_file==True: assert os.path.exists(vocab_file), \"vocab_file does not exist.  Change vocab_from_file to False to create vocab_file.\"\n",
    "        img_folder = os.path.join(cocoapi_loc, 'images/train2014/')\n",
    "        annotations_file = os.path.join(cocoapi_loc, 'annotations/captions_train2014.json')\n",
    "    if mode == 'test':\n",
    "        assert batch_size==1, \"Please change batch_size to 1 if testing your model.\"\n",
    "        assert os.path.exists(vocab_file), \"Must first generate vocab.pkl from training data.\"\n",
    "        assert vocab_from_file==True, \"Change vocab_from_file to True.\"\n",
    "        img_folder = os.path.join(cocoapi_loc, 'images/test2014/')\n",
    "        annotations_file = os.path.join(cocoapi_loc, 'annotations/image_info_test2014.json')\n",
    "    if mode == 'val':\n",
    "        if vocab_from_file==True: assert os.path.exists(vocab_file), \"vocab_file does not exist.  Change vocab_from_file to False to create vocab_file.\"\n",
    "        img_folder = os.path.join(cocoapi_loc, 'images/val2014/')\n",
    "        annotations_file = os.path.join(cocoapi_loc, 'annotations/captions_val2014.json')\n",
    "\n",
    "    # COCO caption dataset.\n",
    "    dataset = CoCoDataset(transform=transform,\n",
    "                          mode=mode,\n",
    "                          batch_size=batch_size,\n",
    "                          vocab_threshold=vocab_threshold,\n",
    "                          vocab_file=vocab_file,\n",
    "                          start_word=start_word,\n",
    "                          end_word=end_word,\n",
    "                          unk_word=unk_word,\n",
    "                          annotations_file=annotations_file,\n",
    "                          vocab_from_file=vocab_from_file,\n",
    "                          img_folder=img_folder)\n",
    "\n",
    "    if mode == 'train' or mode == 'val':\n",
    "        # Randomly sample a caption length, and sample indices with that length.\n",
    "        indices = dataset.get_train_indices()\n",
    "        # Create and assign a batch sampler to retrieve a batch with the sampled indices.\n",
    "        initial_sampler = data.sampler.SubsetRandomSampler(indices=indices)\n",
    "        # data loader for COCO dataset.\n",
    "        data_loader = data.DataLoader(dataset=dataset, \n",
    "                                      num_workers=num_workers,\n",
    "                                      batch_sampler=data.sampler.BatchSampler(sampler=initial_sampler,\n",
    "                                                                              batch_size=dataset.batch_size,\n",
    "                                                                              drop_last=False))\n",
    "    else:\n",
    "        data_loader = data.DataLoader(dataset=dataset,\n",
    "                                      batch_size=dataset.batch_size,\n",
    "                                      shuffle=True,\n",
    "                                      num_workers=num_workers)\n",
    "\n",
    "    return data_loader"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training\n",
    "## Define training dataloader\n",
    "* define a transform to pre-process the training images\n",
    "* initialize a training dataloader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "iFdg6DZmB84N",
    "outputId": "785c814a-55e1-4207-d1f6-cbbf51ff1804"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loading annotations into memory...\n",
      "Done (t=0.64s)\n",
      "creating index...\n",
      "index created!\n",
      "[0/414113] Tokenizing captions...\n",
      "[100000/414113] Tokenizing captions...\n",
      "[200000/414113] Tokenizing captions...\n",
      "[300000/414113] Tokenizing captions...\n",
      "[400000/414113] Tokenizing captions...\n",
      "loading annotations into memory...\n",
      "Done (t=0.64s)\n",
      "creating index...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 785/414113 [00:00<00:52, 7847.97it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index created!\n",
      "Obtaining caption lengths...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 414113/414113 [00:47<00:00, 8725.82it/s]\n"
     ]
    }
   ],
   "source": [
    "# Define a transform to pre-process the training images.\n",
    "transform_train = transforms.Compose([ \n",
    "    transforms.Resize(256),                          # smaller edge of image resized to 256\n",
    "    transforms.RandomCrop(224),                      # get 224x224 crop from random location\n",
    "    transforms.RandomHorizontalFlip(),               # horizontally flip image with probability=0.5\n",
    "    transforms.ToTensor(),                           # convert the PIL Image to a tensor\n",
    "    transforms.Normalize((0.485, 0.456, 0.406),      # normalize image for pre-trained model\n",
    "                         (0.229, 0.224, 0.225))])\n",
    "\n",
    "# Obtain the training loader.\n",
    "train_loader = get_loader(transform=transform_train,\n",
    "                         mode='train',\n",
    "                         batch_size=batch_size,\n",
    "                         vocab_threshold=vocab_threshold,\n",
    "                         vocab_from_file=False)\n",
    "\n",
    "# Store the size of the vocabulary.\n",
    "vocab_size = len(train_loader.dataset.vocab)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Initialize encoder and decoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 117,
     "referenced_widgets": [
      "9a9260a86167496ebba04c1a1def834e",
      "3d12a3efc75342f6a316e57bbb055259",
      "dea161f5239443ad9ff5e9011de6a5db",
      "fc2e24ea8bb54c3183e88e6c77c9ed41",
      "187efbfeefe642a5b6db8a3330faa718",
      "bf5a193c0297497d8833d16de0dfa75c",
      "f56d414fb2ce4cc3b4474a2586deb1bd",
      "8da6aba469b749f2ab8cdf0ae28fd54d"
     ]
    },
    "id": "K-ijOTFDR0aa",
    "outputId": "63a1f9da-8484-4d2b-d608-f24f5241424c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://download.pytorch.org/models/resnet34-333f7ec4.pth\" to /usr4/ec500/brianxu/.cache/torch/hub/checkpoints/resnet34-333f7ec4.pth\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6b5b2eb21eab4b08a40290351f2eb0e7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=87306240.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "EncoderCNN(\n",
       "  (resnet): Sequential(\n",
       "    (0): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
       "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (2): ReLU(inplace=True)\n",
       "    (3): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
       "    (4): Sequential(\n",
       "      (0): BasicBlock(\n",
       "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (1): BasicBlock(\n",
       "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): BasicBlock(\n",
       "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (5): Sequential(\n",
       "      (0): BasicBlock(\n",
       "        (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "          (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): BasicBlock(\n",
       "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): BasicBlock(\n",
       "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (3): BasicBlock(\n",
       "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (6): Sequential(\n",
       "      (0): BasicBlock(\n",
       "        (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): BasicBlock(\n",
       "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): BasicBlock(\n",
       "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (3): BasicBlock(\n",
       "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (4): BasicBlock(\n",
       "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (5): BasicBlock(\n",
       "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (7): Sequential(\n",
       "      (0): BasicBlock(\n",
       "        (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): BasicBlock(\n",
       "        (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "      (2): BasicBlock(\n",
       "        (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (8): AdaptiveAvgPool2d(output_size=(1, 1))\n",
       "  )\n",
       "  (embed): Linear(in_features=512, out_features=300, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Initialize the encoder and move to device\n",
    "encoder = EncoderCNN(embed_size)\n",
    "encoder.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "EnoBqgsNSOw4",
    "outputId": "2c73ea9c-597d-4f97-e2d0-6b8a6fe3ffec"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecoderRNN(\n",
       "  (embedding_layer): Embedding(8852, 300)\n",
       "  (lstm): LSTM(300, 512, batch_first=True)\n",
       "  (linear): Linear(in_features=512, out_features=8852, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Initialize the decoder and move to device\n",
    "decoder = DecoderRNN(embed_size, hidden_size, vocab_size)\n",
    "decoder.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Trainloader/Encoder/Decoder debugging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sampled indices: [407259, 56587, 389309, 205458, 12302, 334916, 81414, 297843, 210846, 400967, 179686, 14060, 349982, 27061, 202118, 280498, 229876, 37620, 87520, 35531, 88198, 178262, 29174, 9201, 210392, 201927, 282855, 186409, 306390, 14915, 262316, 43865, 121337, 39084, 160180, 139822, 16219, 232031, 110777, 238612, 352709, 219227, 19762, 142876, 299199, 240128, 194820, 34637, 99841, 51428, 307230, 64251, 24908, 117041, 39066, 71198, 367042, 365631, 205254, 307919, 409799, 170981, 374363, 144477]\n",
      "images.type: torch.cuda.FloatTensor\n",
      "images.shape: torch.Size([64, 3, 224, 224])\n",
      "captions.type: torch.cuda.LongTensor\n",
      "captions.shape: torch.Size([64, 14])\n"
     ]
    }
   ],
   "source": [
    "# Randomly sample a caption length, and sample indices with that length.\n",
    "indices = train_loader.dataset.get_train_indices()\n",
    "print('sampled indices:', indices)\n",
    "\n",
    "# Create and assign a batch sampler to retrieve a batch with the sampled indices.\n",
    "new_sampler = data.sampler.SubsetRandomSampler(indices=indices)\n",
    "train_loader.batch_sampler.sampler = new_sampler\n",
    "    \n",
    "# Obtain the batch.\n",
    "images, captions = next(iter(train_loader))\n",
    "images = images.to(device)\n",
    "captions = captions.to(device)\n",
    "    \n",
    "print('images.type:',images.type())    \n",
    "print('images.shape:', images.shape)\n",
    "print('captions.type:', captions.type())\n",
    "print('captions.shape:', captions.shape)\n",
    "\n",
    "# (Optional) Uncomment the lines of code below to print the pre-processed images and captions.\n",
    "#print('images:', images)\n",
    "#print('captions:', captions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "type(features): <class 'torch.Tensor'>\n",
      "features.shape: torch.Size([64, 300])\n"
     ]
    }
   ],
   "source": [
    "# Pass the images through the encoder.\n",
    "features = encoder(images)\n",
    "\n",
    "#check feature type and shape\n",
    "print('type(features):', type(features))\n",
    "print('features.shape:', features.shape)\n",
    "\n",
    "assert type(features)==torch.Tensor, \"Encoder output needs to be a PyTorch Tensor.\" \n",
    "assert (features.shape[0]==batch_size) & (features.shape[1]==embed_size), \"The shape of the encoder output is incorrect.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "type(outputs): <class 'torch.Tensor'>\n",
      "outputs.shape: torch.Size([64, 14, 8852])\n",
      "features:  torch.Size([64, 300])\n",
      "captions:  torch.Size([64, 14])\n"
     ]
    }
   ],
   "source": [
    "# Pass the encoder output and captions through the decoder.\n",
    "outputs = decoder(features, captions)\n",
    "\n",
    "print('type(outputs):', type(outputs))\n",
    "print('outputs.shape:', outputs.shape)\n",
    "print(\"features: \", features.shape)\n",
    "print(\"captions: \", captions.shape)\n",
    "\n",
    "assert type(outputs)==torch.Tensor, \"Decoder output needs to be a PyTorch Tensor.\"\n",
    "assert (outputs.shape[0]==batch_size) & (outputs.shape[1]==captions.shape[1]) & (outputs.shape[2]==vocab_size), \"The shape of the decoder output is incorrect.\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define loss, optimizer, steps\n",
    "\n",
    "Loss: Cross Entropy\n",
    "Optimizer: Adam, lr = 0.001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8OUcn19bSekJ",
    "outputId": "2dd3afba-536f-4e99-a80c-e2a40ced28f7"
   },
   "outputs": [],
   "source": [
    "# Define the loss function. \n",
    "criterion = nn.CrossEntropyLoss().cuda() if torch.cuda.is_available() else nn.CrossEntropyLoss()\n",
    "\n",
    "# Specify the learnable parameters of the model.\n",
    "params = list(decoder.parameters()) + list(encoder.embed.parameters())\n",
    "\n",
    "# Define the optimizer.\n",
    "optimizer = torch.optim.Adam(params = params, lr = 0.001)\n",
    "\n",
    "# Set the total number of training steps per epoch.\n",
    "total_step = math.ceil(len(train_loader.dataset.caption_lengths) / train_loader.batch_sampler.batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TRAIN (THIS CELL WILL TAKE A LONG TIME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "dpCY0I3lTQEK",
    "outputId": "74e52363-2e23-460c-bcb8-c0348404b668"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [1/3], Step [100/6471], Loss: 3.8945, Perplexity: 49.1303\n",
      "Epoch [1/3], Step [100/6471], Loss: 3.8945, Perplexity: 49.1303\n",
      "Epoch [1/3], Step [115/6471], Loss: 3.8505, Perplexity: 47.0158"
     ]
    }
   ],
   "source": [
    "# Open the training log file.\n",
    "f = open(log_file, 'w')\n",
    "\n",
    "losses = []\n",
    "perplexities = []\n",
    "\n",
    "for epoch in range(1, num_epochs+1):\n",
    "    \n",
    "    for i_step in range(1, total_step+1):        \n",
    "        # Randomly sample a caption length, and sample indices with that length.\n",
    "        indices = train_loader.dataset.get_train_indices()\n",
    "        \n",
    "        # Create and assign a batch sampler to retrieve a batch with the sampled indices.\n",
    "        new_sampler = data.sampler.SubsetRandomSampler(indices=indices)\n",
    "        train_loader.batch_sampler.sampler = new_sampler\n",
    "        \n",
    "        # Obtain the batch.\n",
    "        images, captions = next(iter(train_loader))\n",
    "\n",
    "        # Move batch of images and captions to GPU if CUDA is available.\n",
    "        images = images.to(device)\n",
    "        captions = captions.to(device)\n",
    "        \n",
    "        # Zero the gradients.\n",
    "        decoder.zero_grad()\n",
    "        encoder.zero_grad()\n",
    "        \n",
    "        # Pass the inputs through the CNN-RNN model.\n",
    "        features = encoder(images)\n",
    "        outputs = decoder(features, captions)\n",
    "        \n",
    "        # Calculate the batch loss.\n",
    "        loss = criterion(outputs.view(-1, vocab_size), captions.view(-1))\n",
    "        if i_step % 100 == 0:\n",
    "            losses.append(loss)\n",
    "            perplexities.append(np.exp(loss.item()))\n",
    "        \n",
    "        # Backward pass.\n",
    "        loss.backward()\n",
    "        \n",
    "        # Update the parameters in the optimizer.\n",
    "        optimizer.step()\n",
    "            \n",
    "        # Get training statistics.\n",
    "        stats = 'Epoch [%d/%d], Step [%d/%d], Loss: %.4f, Perplexity: %5.4f' % (epoch, num_epochs, i_step, total_step, loss.item(), np.exp(loss.item()))\n",
    "        \n",
    "        # Print training statistics (on same line).\n",
    "        print(\"\\r\" + stats, end=\"\")\n",
    "        sys.stdout.flush()\n",
    "        \n",
    "        # Print training statistics to file.\n",
    "        f.write(stats + '\\n')\n",
    "        f.flush()\n",
    "        \n",
    "        # Print training statistics (on different line).\n",
    "        if i_step % print_every == 0:\n",
    "            print(\"\\n\" + stats, end=\"\\n\")\n",
    "            \n",
    "    # Save the weights.\n",
    "    if epoch % save_every == 0:\n",
    "        torch.save(decoder.state_dict(), os.path.join('./models', 'decoder-%d.pkl' % epoch))\n",
    "        torch.save(encoder.state_dict(), os.path.join('./models', 'encoder-%d.pkl' % epoch))\n",
    "\n",
    "# Close the training log file.\n",
    "with open(loss_file, \"wb\") as buttz1:   #Pickling\n",
    "    pickle.dump(losses, buttz1)\n",
    "\n",
    "with open(perplexity_file, \"wb\") as buttz2:   #Pickling\n",
    "    pickle.dump(perplexities, buttz2)\n",
    "\n",
    "f.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot the loss and perplexity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final loss: 2.2123923301696777\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2acf79a716d0>]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "with open(loss_file, 'rb') as ls:\n",
    "                los_ses = pickle.load(ls)\n",
    "print(\"Final loss:\", los_ses[-1].item())\n",
    "ploss=[i.item() for i in los_ses]\n",
    "plt.title('Resnet34 + RNN Loss')\n",
    "plt.plot(ploss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final perplexity: 9.137550302932883\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2aceeb0e2f10>]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "with open(perplexity_file, 'rb') as ps:\n",
    "                perplex_ses = pickle.load(ps)\n",
    "print(\"Final perplexity:\", perplex_ses[-1].item())\n",
    "plex=[i.item() for i in perplex_ses]\n",
    "plt.title('Resnet34 + RNN Perplexities')\n",
    "plt.plot(plex)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Testing\n",
    "\n",
    "## Define testing dataloader\n",
    "* define a transform to pre-process the testing images\n",
    "* initialize a testing dataloader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "6y6OCxMKsJX9",
    "outputId": "8fd8aaee-a229-48e1-82f7-6b4a1c8eeb12"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Vocabulary successfully loaded from vocab.pkl file!\n"
     ]
    }
   ],
   "source": [
    "# Define a transform to pre-process the testing images.\n",
    "transform_test = transforms.Compose([transforms.Resize(256),\n",
    "                                     transforms.CenterCrop(224),\n",
    "                                     transforms.ToTensor(),\n",
    "                                     transforms.Normalize((0.485, 0.456, 0.406),\n",
    "                                                          (0.229, 0.224, 0.225))\n",
    "                                    ])\n",
    "\n",
    "# Create the data loader.\n",
    "test_loader = get_loader(transform=transform_test,    \n",
    "                         mode='test')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load pretrained model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "id": "uwYSz37AzRdl",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecoderRNN(\n",
       "  (embedding_layer): Embedding(8852, 300)\n",
       "  (lstm): LSTM(300, 512, batch_first=True)\n",
       "  (linear): Linear(in_features=512, out_features=8852, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Get vocabulary size\n",
    "vocab_size = len(test_loader.dataset.vocab)\n",
    "\n",
    "# Initialize the encoder and decoder, and set each to inference mode.\n",
    "encoder.eval()\n",
    "decoder = DecoderRNN(embed_size, hidden_size, vocab_size)\n",
    "decoder.eval()\n",
    "\n",
    "# Load the trained weights.\n",
    "encoder.load_state_dict(torch.load(os.path.join('./models', encoder_file)))\n",
    "decoder.load_state_dict(torch.load(os.path.join('./models', decoder_file)))\n",
    "#uncomment this if you aren't using a GPU\n",
    "#encoder.load_state_dict(torch.load(os.path.join('./models', encoder_file),map_location=torch.device('cpu')))\n",
    "#decoder.load_state_dict(torch.load(os.path.join('./models', decoder_file),map_location=torch.device('cpu')))\n",
    "\n",
    "encoder.to(device)\n",
    "decoder.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Display a single image + caption prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_prediction(encoder, decoder, test_loader):\n",
    "    orig_image, image = next(iter(test_loader))\n",
    "    #print('orig_image ' , orig_image.shape)\n",
    "    print('image: ' , image.shape)\n",
    "    plt.imshow(np.squeeze(orig_image))\n",
    "    plt.title('Sample Image')\n",
    "    plt.show()\n",
    "    image = image.to(device)\n",
    "    features = encoder(image).unsqueeze(1)\n",
    "    output = decoder.sample(features)    \n",
    "    sentence = clean_sentence(output, test_loader)\n",
    "    print(sentence)\n",
    "\n",
    "def clean_sentence(output, data_loader):\n",
    "    words_sequence = []\n",
    "    for i in output:\n",
    "        if (i == 1):\n",
    "            continue\n",
    "        words_sequence.append(data_loader.dataset.vocab.idx2word[i])\n",
    "    words_sequence = words_sequence[1:-1] \n",
    "    sentence = ' '.join(words_sequence) \n",
    "    sentence = sentence.capitalize()    \n",
    "    return sentence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 298
    },
    "id": "ww8VWlnr0Wz5",
    "outputId": "e0ff6eed-f10e-4b9a-9f92-2e589e43592a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image:  torch.Size([1, 3, 224, 224])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A herd of cattle standing on top of a lush green hillside . .\n"
     ]
    }
   ],
   "source": [
    "get_prediction(encoder, decoder, test_loader)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate BLEU score\n",
    "Calculate the BLEU score of the model over the entire dataset by averaging the BLEU scores over several batches.\n",
    "\n",
    "### Define a data loader\n",
    "This data loader should return images and their corresponding ground-truth captions from the validation dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Vocabulary successfully loaded from vocab.pkl file!\n",
      "loading annotations into memory...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 855/202654 [00:00<00:23, 8540.30it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done (t=0.30s)\n",
      "creating index...\n",
      "index created!\n",
      "Obtaining caption lengths...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 202654/202654 [00:23<00:00, 8721.73it/s]\n"
     ]
    }
   ],
   "source": [
    "transform_val = transforms.Compose([transforms.Resize(256),\n",
    "                                     transforms.CenterCrop(224),\n",
    "                                     transforms.ToTensor(),\n",
    "                                     transforms.Normalize((0.485, 0.456, 0.406),\n",
    "                                                          (0.229, 0.224, 0.225))\n",
    "                                    ])\n",
    "val_loader = get_loader(transform=transform_val,\n",
    "                         mode='val',batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "SyntaxError",
     "evalue": "unexpected character after line continuation character (<ipython-input-1-7255e203fc92>, line 40)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;36m  File \u001b[0;32m\"<ipython-input-1-7255e203fc92>\"\u001b[0;36m, line \u001b[0;32m40\u001b[0m\n\u001b[0;31m    \\4oader))\u001b[0m\n\u001b[0m             ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m unexpected character after line continuation character\n"
     ]
    }
   ],
   "source": [
    "def clean_sentence_bleu(output, data_loader):\n",
    "    words_sequence = []\n",
    "    for i in output:\n",
    "        i = int(i)\n",
    "        if (i == 1):\n",
    "            continue\n",
    "        words_sequence.append(data_loader.dataset.vocab.idx2word[i])\n",
    "    words_sequence = words_sequence[1:-1] \n",
    "#   sentence = ' '.join(words_sequence) \n",
    "#   sentence = sentence.capitalize()    \n",
    "    return words_sequence\n",
    "\n",
    "def get_avg_bleu(encoder, decoder, dataloader, device):\n",
    "    '''Determine the average BLEU score across the entire dataset\n",
    "    '''\n",
    "    with torch.no_grad():\n",
    "        total_score1 = 0\n",
    "        total_score2 = 0\n",
    "        total_score3 = 0\n",
    "        total_score4 = 0\n",
    "        total_num = 0\n",
    "        for data in tqdm(dataloader):\n",
    "            torch.cuda.empty_cache()\n",
    "            \n",
    "            #load images and reference captions from the dataset\n",
    "            images, captions_ref = data\n",
    "            images = images.to(device)\n",
    "            print('captions_ref: ',captions_ref.type())\n",
    "            # Get caption predictions for the current batch of validation images\n",
    "            captions_cand = get_batch_candidates(encoder, decoder, images)\n",
    "            \n",
    "            #calculate the bleu score for the entire batch\n",
    "            print(captions_ref.shape)\n",
    "            refs = []\n",
    "            cands = []\n",
    "            for captions in captions_ref:\n",
    "                #captions = captions.float()\n",
    "                #print(captions.type())\n",
    "                refs.append(clean_sentence_bleu(captions, datal+++++++++++++++++9\n",
    "                                                \\4oader))\n",
    "                \n",
    "            for captions_2 in captions_cand:\n",
    "                cands.append(clean_sentence_bleu(captions_2, dataloader))\n",
    "#             print(refs.size())\n",
    "#             print(cands.size())\n",
    "            #batch_score = get_batch_bleu(captions_ref, captions_cand)\n",
    "            batch_score1, batch_score2, batch_score3, batch_score4 = get_batch_bleu(refs, cands)\n",
    "            #increment total_score and total_num\n",
    "            total_score1 = total_score1 + batch_score1\n",
    "            total_score2 = total_score2 + batch_score2\n",
    "            total_score3 = total_score3 + batch_score3\n",
    "            total_score4 = total_score4 + batch_score4\n",
    "            \n",
    "            total_num = total_num + dataloader.dataset.batch_size\n",
    "        \n",
    "            \n",
    "            #print('Total Num: ',total_num)\n",
    "            \n",
    "        avg_score1 = total_score1/total_num\n",
    "        avg_score2 = total_score2/total_num\n",
    "        avg_score3 = total_score3/total_num\n",
    "        avg_score4 = total_score4/total_num\n",
    "        \n",
    "        return avg_score1, avg_score2, avg_score3, avg_score4\n",
    "\n",
    "def get_batch_candidates(encoder, decoder, images):\n",
    "    '''Gets predictions for the current batch of images\n",
    "    '''\n",
    "    with torch.no_grad():\n",
    "        captions_cand = []\n",
    "\n",
    "        features_batch = encoder(images).unsqueeze(1)\n",
    "\n",
    "        for features in features_batch:\n",
    "            new_cand = decoder.sample(features.unsqueeze(0))\n",
    "            captions_cand.append(new_cand)\n",
    "\n",
    "        return torch.Tensor(captions_cand)\n",
    "\n",
    "def get_batch_bleu(captions_ref, captions_cand):\n",
    "    '''Compute the total BLEU score over elements in a batch\n",
    "    '''\n",
    "    with torch.no_grad():\n",
    "        scores1 = 0\n",
    "        scores2 = 0\n",
    "        scores3 = 0\n",
    "        scores4 = 0\n",
    "        for i in range(len(captions_ref)):\n",
    "            #print(len(ref))\n",
    "            captions_cand[i] = [x for x in captions_cand[i] if x != '.']\n",
    "            print('captions_ref= ',captions_ref[i])\n",
    "            print('captions_cand= ',captions_cand[i])\n",
    "            scores1 += nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i], weights=(1, 0, 0, 0))\n",
    "            scores2 += nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i], weights=(0.5, 0.5, 0, 0))\n",
    "            scores3 += nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i], weights=(0.33, 0.33, 0.33, 0.33))\n",
    "            scores4 += nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i], weights=(0.25, 0.25, 0.25, 0.25))\n",
    "            print('Score : ',nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i],weights=(1, 0, 0, 0)))\n",
    "            print('Score : ',nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i],weights=(0.5, 0.5, 0, 0)))\n",
    "            print('Score : ',nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i],weights=(0.33, 0.33, 0.33, 0)))\n",
    "            print('Score : ',nltk.translate.bleu_score.sentence_bleu([captions_ref[i]], captions_cand[i],weights=(0.25, 0.25, 0.25, 0.25)))\n",
    "        return scores1, scores2, scores3, scores4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'get_avg_bleu' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-2-27182d6c449b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mbleu_score1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score4\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_avg_bleu\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mencoder\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdecoder\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_loader\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\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      2\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mbleu_score1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbleu_score4\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;31mNameError\u001b[0m: name 'get_avg_bleu' is not defined"
     ]
    }
   ],
   "source": [
    "bleu_score1, bleu_score2, bleu_score3, bleu_score4 = get_avg_bleu(encoder, decoder, val_loader, device)\n",
    "print([bleu_score1, bleu_score2, bleu_score3, bleu_score4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\0[[[[[[[+;3]]]]]]]"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "Model0.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "187efbfeefe642a5b6db8a3330faa718": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": "initial"
     }
    },
    "3d12a3efc75342f6a316e57bbb055259": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "8da6aba469b749f2ab8cdf0ae28fd54d": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "9a9260a86167496ebba04c1a1def834e": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_dea161f5239443ad9ff5e9011de6a5db",
       "IPY_MODEL_fc2e24ea8bb54c3183e88e6c77c9ed41"
      ],
      "layout": "IPY_MODEL_3d12a3efc75342f6a316e57bbb055259"
     }
    },
    "bf5a193c0297497d8833d16de0dfa75c": {
     "model_module": "@jupyter-widgets/base",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "dea161f5239443ad9ff5e9011de6a5db": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "100%",
      "description_tooltip": null,
      "layout": "IPY_MODEL_bf5a193c0297497d8833d16de0dfa75c",
      "max": 102502400,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_187efbfeefe642a5b6db8a3330faa718",
      "value": 102502400
     }
    },
    "f56d414fb2ce4cc3b4474a2586deb1bd": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "fc2e24ea8bb54c3183e88e6c77c9ed41": {
     "model_module": "@jupyter-widgets/controls",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_8da6aba469b749f2ab8cdf0ae28fd54d",
      "placeholder": "​",
      "style": "IPY_MODEL_f56d414fb2ce4cc3b4474a2586deb1bd",
      "value": " 97.8M/97.8M [06:02&lt;00:00, 283kB/s]"
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}