pix2code - Pytorch.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"pix2code - Pytorch.ipynb","provenance":[{"file_id":"1Kvqns6sDdUUVQhIYXzq9yanO9lNzv2iR","timestamp":1617576169457}],"collapsed_sections":[],"toc_visible":true,"machine_shape":"hm"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"code","metadata":{"id":"IEjCiu0H8Rno","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1618194667040,"user_tz":240,"elapsed":1364,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"d95594f6-93db-4433-c30e-88e2d5d1d9b4"},"source":["!nvidia-smi"],"execution_count":1,"outputs":[{"output_type":"stream","text":["Mon Apr 12 02:31:07 2021       \n","+-----------------------------------------------------------------------------+\n","| NVIDIA-SMI 460.67       Driver Version: 460.32.03    CUDA Version: 11.2     |\n","|-------------------------------+----------------------+----------------------+\n","| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n","| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n","|                               |                      |               MIG M. |\n","|===============================+======================+======================|\n","|   0  Tesla V100-SXM2...  Off  | 00000000:00:04.0 Off |                    0 |\n","| N/A   41C    P0    26W / 300W |      0MiB / 16160MiB |      0%      Default |\n","|                               |                      |                  N/A |\n","+-------------------------------+----------------------+----------------------+\n","                                                                               \n","+-----------------------------------------------------------------------------+\n","| Processes:                                                                  |\n","|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n","|        ID   ID                                                   Usage      |\n","|=============================================================================|\n","|  No running processes found                                                 |\n","+-----------------------------------------------------------------------------+\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"ttHg96xT8nsu","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1618194686352,"user_tz":240,"elapsed":17220,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"21119d06-e87a-4fdb-d457-07daa9082b74"},"source":["from google.colab import drive\n","drive.mount('/content/drive')\n","%cd \"/content/drive/My Drive/pytorch_pix2code\"\n"],"execution_count":2,"outputs":[{"output_type":"stream","text":["Mounted at /content/drive\n","/content/drive/My Drive/pytorch_pix2code\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"dP2NfBQ5V7-W","executionInfo":{"status":"ok","timestamp":1618169077735,"user_tz":240,"elapsed":498,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"01a9ab71-f066-4348-81f6-c7d3905d47c7"},"source":["%cd \"/content/drive/My Drive/pytorch_pix2code\""],"execution_count":null,"outputs":[{"output_type":"stream","text":["/content/drive/My Drive/pytorch_pix2code\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"yDBILtSCD8FN"},"source":["# Only run it if you don't have the repo\n","#!git clone https://github.com/VaibhavYadav/pytorch_pix2code.git #might change it to our github repo?"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"N4vIeJIkcmsl"},"source":["# ADJUSTING DATASET\n","no need to run this part unless\n","you manipulating the datasets\n","RUN ONLY ONCE"]},{"cell_type":"markdown","metadata":{"id":"FS40fyRpVkai"},"source":["Running Code below to separate new dataset to training and evaluate dataset\n","\n"]},{"cell_type":"code","metadata":{"id":"dtOQ78sQPi0D"},"source":["\"\"\"\n","%cd \"/content/drive/My Drive\"\n","\n","\n","import os\n","import numpy as np\n","from shutil import copyfile\n","\n","\n","input_path = './sketch2code/data'\n","output_path = 'pytorch_pix2code/sketch_dataset/'\n","eval_split_percent = 0.10\n","\n","\n","# List of every datapoint filename\n","paths = []\n","for f in os.listdir(input_path):\n","    if f.find('.gui') != -1:\n","        file_name = f[:f.find('.gui')]\n","        if os.path.isfile('{}/{}.png'.format(input_path, file_name)):\n","            paths.append(file_name)\n","\n","# Split the data in training and evaluation set\n","eval_sample_number = int(len(paths) * eval_split_percent)\n","np.random.shuffle(paths)\n","eval_set = paths[:eval_sample_number]\n","train_set = paths[eval_sample_number:]\n","\n","for path in eval_set:\n","    copyfile('{}/{}.png'.format(input_path, path), '{}/{}/{}.png'.format(os.path.dirname(output_path), 'evaluation', path))\n","    copyfile('{}/{}.gui'.format(input_path, path), '{}/{}/{}.gui'.format(os.path.dirname(output_path), 'evaluation', path))\n","\n","for path in train_set:\n","    copyfile('{}/{}.png'.format(input_path, path), '{}/{}/{}.png'.format(os.path.dirname(output_path), 'training', path))\n","    copyfile('{}/{}.gui'.format(input_path, path), '{}/{}/{}.gui'.format(os.path.dirname(output_path), 'training', path))\n"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"Of6wt98H54bq"},"source":["converting some sketch images to 4 channels (only need to be done once)"]},{"cell_type":"code","metadata":{"id":"kN7cOPN3591X"},"source":["'''\n","input_path = \"/content/drive/My Drive/pytorch_pix2code/sketch_dataset/training\"\n","output_path = \"/content/drive/My Drive/pytorch_pix2code/sketch_dataset2/training\"\n","\n","import glob\n","def png2pnga(path_open, path_save, files):\n","    for name in glob.glob(path_open + files): \n","        if name.find('.png') != 1:\n","            im = Image.open(name).convert('RGBA')\n","            im.save(path_save + name[len(path_open):], \"PNG\", quality=100)\n","png2pnga(input_path, output_path, \"/*.png\")\n","'''"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"t0q-LX8S70LT"},"source":["Moving Gui to new destination"]},{"cell_type":"code","metadata":{"id":"SVc3jsHt7gnl"},"source":["'''\n","from shutil import copyfile\n","import os\n","\n","paths = []\n","for f in os.listdir(input_path):\n","    if f.find('.gui') != -1:\n","        file_name = f[:f.find('.gui')]\n","        if os.path.isfile('{}/{}.png'.format(input_path, file_name)):\n","            paths.append(file_name)\n","for path in paths:\n","    print(path)\n","    copyfile('{}/{}.gui'.format(input_path, path), '{}/{}.gui'.format(output_path, path))\n","'''"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"y66ZQvQqGDPG"},"source":["Deleting some Gui files in the Desired Folder"]},{"cell_type":"code","metadata":{"id":"LpSpthYiKHQG"},"source":["\"\"\"\n","from shutil import copyfile\n","import os\n","input_path = r\"/content/drive/My Drive/pytorch_pix2code/sketch_dataset2/evaluation\"\n","paths = []\n","for f in os.listdir(input_path):\n","    if f.find('.gui') != -1:\n","        os.remove(f\"{input_path}/{f}\")\n","\"\"\"\n"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"HyCgFh38EJrQ"},"source":["# IMPORTING LIBRARIES"]},{"cell_type":"code","metadata":{"id":"Xs62SAj396wv","executionInfo":{"status":"ok","timestamp":1618194692090,"user_tz":240,"elapsed":4927,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}}},"source":["from util import UIDataset, Vocabulary, SketchDataset, SketchTest, UITest\n","import torch.nn as nn\n","import torch.nn.functional as F\n","import numpy as np\n","import torch\n","from torch.utils.data import DataLoader\n","from model import *\n","#from sketch_model import *\n","from torchvision import transforms\n","from PIL import Image"],"execution_count":3,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"IpG3QUaeEgr2"},"source":["# LOADING DATASET"]},{"cell_type":"code","metadata":{"id":"HPszw-zN_PKW","executionInfo":{"status":"ok","timestamp":1618194895543,"user_tz":240,"elapsed":900,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}}},"source":["# dataset = UIDataset('./sketch_dataset/training', 'voc.pkl') # UIDataset is used for screenshots\n","dataset = SketchDataset('./sketch_dataset_new/training', './compiler/vocab.pkl') # SketchDataset is used for Sketches (.png only for now)"],"execution_count":6,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"8JwEY3l2RuN4"},"source":["Sanity Check Test"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":291},"id":"jWdp4S2YQoIK","executionInfo":{"status":"ok","timestamp":1618195840451,"user_tz":240,"elapsed":748,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"11da5326-11ad-478d-fe73-d0fb8fa71c1f"},"source":["import cv2\n","#Shuffling the Dataset\n","loader = DataLoader(dataset,batch_size=75, shuffle=True) \n","image, *_ = dataset.__getitem__(9)\n","print(image.shape)\n","t = transforms.ToPILImage()\n","image = image.unsqueeze(0)\n","t(image.squeeze())\n","\n"],"execution_count":44,"outputs":[{"output_type":"stream","text":["torch.Size([3, 256, 256])\n"],"name":"stdout"},{"output_type":"execute_result","data":{"image/png":"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\n","text/plain":["<PIL.Image.Image image mode=RGB size=256x256 at 0x7F08FA0F4E10>"]},"metadata":{"tags":[]},"execution_count":44}]},{"cell_type":"markdown","metadata":{"id":"Pm6eW7AQIIPf"},"source":["# TRAINING"]},{"cell_type":"markdown","metadata":{"id":"kQoJ0raIILti"},"source":["Hyper Parameters & defining Model"]},{"cell_type":"code","metadata":{"id":"kh2104E-G-PP"},"source":["input_size = 28\n","model = Pix2Code(input_size).cuda()\n","loss_fn = nn.CrossEntropyLoss()\n","optimizer = torch.optim.Adam(model.parameters(), lr = 0.0001)\n","epochs = 10\n","model_output_name = 'Sketch2Code_v4'"],"execution_count":null,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"PzvPr7mgI-XJ"},"source":["Training"]},{"cell_type":"code","metadata":{"id":"1WV84ehvISCx","colab":{"base_uri":"https://localhost:8080/","height":579},"executionInfo":{"status":"error","timestamp":1618195783586,"user_tz":240,"elapsed":117749,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"54a95041-8dab-4601-c462-009f45b8f2f4"},"source":["for epoch in range(epochs):\n","    model.zero_grad()\n","\n","    for j, data in enumerate(loader):\n","        # Extracting Dataset \n","        image, context, expected_output = data   \n","\n","        #image = image.unsqueeze(0).cuda()\n","        #context = context.unsqueeze(0).cuda()\n","        #image = image.cuda()\n","        #context = context.cuda()\n","        expected_output = expected_output.cuda()\n","        # Forward Propagation, Loss, and Back-Propagation\n","        output = model(image, context)\n","        output = output.squeeze(0)\n","        expected_output = torch.argmax(expected_output, 1)\n","        \n","        loss = loss_fn(output, expected_output)\n","        loss.backward()\n","\n","        # Updating weights every 10 steps, and resetting gradients (since it accumulates in Pytorch)\n","        if j%10 == 0:\n","            optimizer.step()\n","            print('Loss: {}, Epoch: {}'.format(loss.data, epoch))\n","            model.zero_grad()\n","\n","    torch.save(model.state_dict(), f\"./model_weights/{model_output_name}.weights\")"],"execution_count":40,"outputs":[{"output_type":"stream","text":["Loss: 3.328197956085205, Epoch: 0\n","Loss: 3.2559142112731934, Epoch: 0\n","Loss: 3.1718904972076416, Epoch: 0\n","Loss: 3.022108316421509, Epoch: 0\n","Loss: 2.8318932056427, Epoch: 0\n","Loss: 2.6593616008758545, Epoch: 0\n","Loss: 2.4772450923919678, Epoch: 0\n","Loss: 2.3386759757995605, Epoch: 0\n","Loss: 2.266557216644287, Epoch: 0\n","Loss: 2.2563796043395996, Epoch: 0\n","Loss: 2.1964049339294434, Epoch: 0\n"],"name":"stdout"},{"output_type":"error","ename":"KeyboardInterrupt","evalue":"ignored","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)","\u001b[0;32m<ipython-input-40-886538a7448d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m     \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mzero_grad\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m     \u001b[0;32mfor\u001b[0m \u001b[0mj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdataset\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[0m\u001b[1;32m      5\u001b[0m         \u001b[0;31m# Extracting Dataset\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m         \u001b[0mimage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexpected_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/My Drive/pytorch_pix2code/util.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, index)\u001b[0m\n\u001b[1;32m    135\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    136\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__getitem__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 137\u001b[0;31m         \u001b[0mimage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mresize_img\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfloat\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    138\u001b[0m         \u001b[0mcontext\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprediction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread_gui\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'{}/{}.gui'\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfile_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpaths\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    139\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mimage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprediction\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/content/drive/My Drive/pytorch_pix2code/util.py\u001b[0m in \u001b[0;36mresize_img\u001b[0;34m(self, index)\u001b[0m\n\u001b[1;32m    140\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    141\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mresize_img\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 142\u001b[0;31m         \u001b[0mimg_rgb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcv2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34mf\"{self.file_path}/{self.paths[index]}.png\"\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    143\u001b[0m         \u001b[0mimg_grey\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcv2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcvtColor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg_rgb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcv2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mCOLOR_BGR2GRAY\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    144\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minverse\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;31mKeyboardInterrupt\u001b[0m: "]}]},{"cell_type":"markdown","metadata":{"id":"mNOGw1XCI8H4"},"source":["# TESTING"]},{"cell_type":"code","metadata":{"id":"NAxFnLyXIkMp","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1618190937664,"user_tz":240,"elapsed":1677,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"a68e67fc-1629-40c8-cd4d-81fb7abf5d12"},"source":["input_size = 28\n","model = Pix2Code(input_size)\n","model.load_state_dict(torch.load(f\"./model_weights/{model_output_name}.weights\"))\n","model.cuda().eval()"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Pix2Code(\n","  (image_encoder): ImageEncoder(\n","    (conv1): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1))\n","    (conv2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1))\n","    (conv3): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1))\n","    (conv4): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1))\n","    (conv5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1))\n","    (conv6): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1))\n","    (fc1): Linear(in_features=100352, out_features=1024, bias=True)\n","    (fc2): Linear(in_features=1024, out_features=1024, bias=True)\n","  )\n","  (context_encoder): ContextEncoder(\n","    (rnn): RNN(28, 128, num_layers=2, batch_first=True)\n","  )\n","  (decoder): Decoder(\n","    (rnn): RNN(1152, 512, num_layers=2, batch_first=True)\n","    (l1): Linear(in_features=512, out_features=28, bias=True)\n","  )\n",")"]},"metadata":{"tags":[]},"execution_count":8}]},{"cell_type":"code","metadata":{"id":"ukP7bybdJeIh"},"source":["#test_data = UIDataset('./sketch_dataset/evaluation', 'voc.pkl')\n","#test_data = SketchTest('./validation', 'voc.pkl')\n","test_data = SketchTest('./sketch_dataset_new/training', './compiler/vocab.pkl')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":185},"id":"8Td62-U9xPRq","executionInfo":{"status":"error","timestamp":1618177927433,"user_tz":240,"elapsed":609,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"5a39ce21-521c-4e74-ffec-0fd224a40484"},"source":[""],"execution_count":null,"outputs":[{"output_type":"error","ename":"ValueError","evalue":"ignored","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)","\u001b[0;32m<ipython-input-75-69c86cef9b21>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mtest_data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mimage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexpected_output\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtest_data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m","\u001b[0;31mValueError\u001b[0m: not enough values to unpack (expected 3, got 0)"]}]},{"cell_type":"markdown","metadata":{"id":"BlJDEgOIil6C"},"source":["Just to visualize our current tokens"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"zeuuUvYutLzk","executionInfo":{"status":"ok","timestamp":1618192180625,"user_tz":240,"elapsed":510,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"8662bbe0-084a-472f-af85-864309f34a84"},"source":["vocab = Vocabulary('./compiler/vocab.pkl')\n","vocab.index_to_vocab"],"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["{0: '\\n',\n"," 1: ' ',\n"," 2: ',',\n"," 3: '<END>',\n"," 4: '<START>',\n"," 5: 'btn-active',\n"," 6: 'btn-green',\n"," 7: 'btn-inactive',\n"," 8: 'btn-orange',\n"," 9: 'btn-red',\n"," 10: 'double',\n"," 11: 'header',\n"," 12: 'quadruple',\n"," 13: 'row',\n"," 14: 'single',\n"," 15: 'small-title',\n"," 16: 'text',\n"," 17: '{',\n"," 18: '}',\n"," 19: 'css-file-name',\n"," 20: 'header-right',\n"," 21: 'big-title',\n"," 22: 'big-text',\n"," 23: 'small-text',\n"," 24: 'img',\n"," 25: 'empty',\n"," 26: 'container',\n"," 27: 'btn'}"]},"metadata":{"tags":[]},"execution_count":44}]},{"cell_type":"markdown","metadata":{"id":"ASBMVhXAO5ky"},"source":["Visualizing the Preprocessed & Processed input image"]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/","height":291},"id":"MX3N7ePyJekL","executionInfo":{"status":"ok","timestamp":1618192344427,"user_tz":240,"elapsed":537,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"d139f98d-7f1a-4c67-9ccd-39b9a9807a5d"},"source":["output_name = f\"{model_output_name}_Test_1\" #Output Directory\n","index = np.random.randint(len(test_data))\n","#index = -1\n","image = test_data.__getitem__(index)\n","print(test_data.paths[index])\n","t = transforms.ToPILImage()\n","image = image.unsqueeze(0)\n","t(image.squeeze())\n","# unsqueezeing to insert into predictor"],"execution_count":null,"outputs":[{"output_type":"stream","text":["0180\n"],"name":"stdout"},{"output_type":"execute_result","data":{"image/png":"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\n","text/plain":["<PIL.Image.Image image mode=RGB size=256x256 at 0x7F443C353F50>"]},"metadata":{"tags":[]},"execution_count":47}]},{"cell_type":"code","metadata":{"id":"aasmGY6IxI5w"},"source":[""],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"NyVBHDhJJh9b","executionInfo":{"status":"ok","timestamp":1618192350812,"user_tz":240,"elapsed":875,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"53967af0-341c-4263-b535-544799ac5e70"},"source":["image = image.cuda()\n","ct = [] #ct stands for context, and we put \" \" and \"<START>\" as our first input to trigger LSTM based on CNN\n","# Initializing context\n","ct.append(vocab.to_vec(' '))\n","ct.append(vocab.to_vec('<START>'))\n","output = ''   #This is the.gui file, doesn't need HTML code\n","\n","# PREDICTING\n","for i in range(200):\n","    context = torch.tensor(ct).unsqueeze(0).float().cuda()\n","    index = torch.argmax(model(image, context), 2).squeeze()[-1:].squeeze()\n","    v = vocab.to_vocab(int(index))\n","    if v == '<END>':\n","        break\n","    output += v\n","    ct.append(vocab.to_vec(v))\n","\n","with open(f\"./compiler/{output_name}.gui\", 'w') as f:\n","    f.write(output)\n","print(\" \")\n","print(\"GUI OUTPUT: \")\n","print(output)"],"execution_count":null,"outputs":[{"output_type":"stream","text":[" \n","GUI OUTPUT: \n","header{\n","quadruple{\n","btn-active\n","}\n","double{\n","btn-inactive\n","}\n","quadruple{\n","btn-active\n","}\n","}\n","container{\n","row{\n","single{\n","empty\n","}\n","}\n","row{\n","single{\n","big-text,text\n","}\n","}\n","}\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"IshAqktjN3ZQ","executionInfo":{"status":"ok","timestamp":1618192024807,"user_tz":240,"elapsed":1035,"user":{"displayName":"ous Abou Ras","photoUrl":"","userId":"00588357774699055147"}},"outputId":"38d3fded-45cc-4e97-bd73-9994bb177764"},"source":["%cd compiler\n","!python web-compiler.py output/{output_name}.gui\n","%cd \"/content/drive/My Drive/pytorch_pix2code\""],"execution_count":null,"outputs":[{"output_type":"stream","text":["/content/drive/My Drive/pytorch_pix2code/compiler\n","/content/drive/My Drive/pytorch_pix2code\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"4ZBH9m9uOirL"},"source":[""],"execution_count":null,"outputs":[]}]}