nb/cvae/LabeledConvCVAE.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Labeled Convolutional CVAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "19d0f047",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torchvision import transforms\n",
    "from torch.utils.data import DataLoader, SubsetRandomSampler\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load pallets library from project root\n",
    "import sys, os\n",
    "sys.path.append(os.path.abspath(os.path.join(os.getcwd(), '..', '..')))\n",
    "\n",
    "from pallets import (\n",
    "    images as I,\n",
    "    datasets as DS,\n",
    "    models as M,\n",
    "    logging as L\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "844bbd48",
   "metadata": {},
   "source": [
    "# Settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9b8ec428",
   "metadata": {},
   "outputs": [],
   "source": [
    "SAVE_NAME = 'cvae.conv.labeled'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "ae1b5c2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "USE_GPU = True\n",
    "LOG_LEVEL = 'INFO'\n",
    "LABELS_FILE = 'pallets_labels.json'\n",
    "\n",
    "TEST_SIZE = 1000\n",
    "EPOCHS = 100\n",
    "LR = 3e-04\n",
    "BATCH_SIZE = 32\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Logger pallets (INFO)>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# To GPU, or not to GPU\n",
    "device = M.get_device(require_gpu=USE_GPU)\n",
    "\n",
    "# Logging\n",
    "L.init_logger(level=LOG_LEVEL, notebook=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "35fef1c0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO | loading punk images\n",
      "INFO | - image 0\n",
      "INFO | - image 1000\n",
      "INFO | - image 2000\n",
      "INFO | - image 3000\n",
      "INFO | - image 4000\n",
      "INFO | - image 5000\n",
      "INFO | - image 6000\n",
      "INFO | - image 7000\n",
      "INFO | - image 8000\n",
      "INFO | - image 9000\n",
      "INFO | punk images complete\n",
      "INFO | loading punk labels: pallets_labels.json\n"
     ]
    }
   ],
   "source": [
    "all_colors = I.get_punk_colors()\n",
    "mapper = DS.ColorOneHotMapper(all_colors)\n",
    "dataset = DS.OneHotCPunksDataset(\n",
    "    mapper, labels_file=LABELS_FILE, test_size=TEST_SIZE\n",
    ")\n",
    "\n",
    "# Uncomment to put entire dataset in GPU RAM\n",
    "# dataset = DS.FastOneHotCPunksDataset(\n",
    "#     device, mapper, labels_file=LABELS_FILE, test_size=TEST_SIZE\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_sampler = SubsetRandomSampler(dataset.train_idx)\n",
    "test_sampler = SubsetRandomSampler(dataset.test_idx)\n",
    "\n",
    "train_loader = DataLoader(dataset, batch_size=BATCH_SIZE, sampler=train_sampler)\n",
    "test_loader = DataLoader(dataset, batch_size=BATCH_SIZE, sampler=test_sampler)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b39e0222",
   "metadata": {},
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7e4d5fd0",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_dim = 222\n",
    "latent_dim = 32\n",
    "classes_dim = len(dataset._label_keys)  # 92"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9521e7fe",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO | model: pallets.models.cvae.LabeledConvCVAE\n",
      "INFO | criterion: pallets.models.cvae.Loss\n",
      "INFO | learn rate: 0.0003\n",
      "INFO | epochs: 100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO | epoch 1 (  0%) loss: 2005726.250000\n",
      "INFO | epoch 1 ( 35%) loss: 339873.233601\n",
      "INFO | epoch 1 ( 70%) loss: 241921.033271\n",
      "INFO | epoch 1 (100%) loss: 212578.157303\n",
      "INFO | epoch 1 (test) loss: 135194.287170\n",
      "INFO | epoch 2 (  0%) loss: 133063.625000\n",
      "INFO | epoch 2 ( 35%) loss: 135468.649134\n",
      "INFO | epoch 2 ( 70%) loss: 134701.380869\n",
      "INFO | epoch 2 (100%) loss: 133864.933213\n",
      "INFO | epoch 2 (test) loss: 129977.242432\n",
      "INFO | epoch 3 (  0%) loss: 131311.171875\n",
      "INFO | epoch 3 ( 35%) loss: 125220.036819\n",
      "INFO | epoch 3 ( 70%) loss: 118753.395678\n",
      "INFO | epoch 3 (100%) loss: 115943.919679\n",
      "INFO | epoch 3 (test) loss: 107423.998474\n",
      "INFO | epoch 4 (  0%) loss: 111806.335938\n",
      "INFO | epoch 4 ( 35%) loss: 108740.759282\n",
      "INFO | epoch 4 ( 70%) loss: 107635.042366\n",
      "INFO | epoch 4 (100%) loss: 106917.455307\n",
      "INFO | epoch 4 (test) loss: 104194.661621\n",
      "INFO | epoch 5 (  0%) loss: 107273.109375\n",
      "INFO | epoch 5 ( 35%) loss: 104704.202661\n",
      "INFO | epoch 5 ( 70%) loss: 104120.509950\n",
      "INFO | epoch 5 (100%) loss: 103120.449918\n",
      "INFO | epoch 5 (test) loss: 100183.707458\n",
      "INFO | epoch 6 (  0%) loss: 100755.328125\n",
      "INFO | epoch 6 ( 35%) loss: 98742.478342\n",
      "INFO | epoch 6 ( 70%) loss: 97778.821556\n",
      "INFO | epoch 6 (100%) loss: 96094.667990\n",
      "INFO | epoch 6 (test) loss: 89667.205017\n",
      "INFO | epoch 7 (  0%) loss: 94161.320312\n",
      "INFO | epoch 7 ( 35%) loss: 89924.393410\n",
      "INFO | epoch 7 ( 70%) loss: 88600.249495\n",
      "INFO | epoch 7 (100%) loss: 86973.379246\n",
      "INFO | epoch 7 (test) loss: 81233.247681\n",
      "INFO | epoch 8 (  0%) loss: 76392.257812\n",
      "INFO | epoch 8 ( 35%) loss: 80204.200418\n",
      "INFO | epoch 8 ( 70%) loss: 77591.933982\n",
      "INFO | epoch 8 (100%) loss: 75072.277223\n",
      "INFO | epoch 8 (test) loss: 66583.268372\n",
      "INFO | epoch 9 (  0%) loss: 66668.890625\n",
      "INFO | epoch 9 ( 35%) loss: 65249.233718\n",
      "INFO | epoch 9 ( 70%) loss: 63009.927958\n",
      "INFO | epoch 9 (100%) loss: 61152.793886\n",
      "INFO | epoch 9 (test) loss: 55274.693268\n",
      "INFO | epoch 10 (  0%) loss: 50059.707031\n",
      "INFO | epoch 10 ( 35%) loss: 51894.150371\n",
      "INFO | epoch 10 ( 70%) loss: 50475.714319\n",
      "INFO | epoch 10 (100%) loss: 49160.190087\n",
      "INFO | epoch 10 (test) loss: 47870.675446\n",
      "INFO | epoch 11 (  0%) loss: 49367.792969\n",
      "INFO | epoch 11 ( 35%) loss: 43010.778639\n",
      "INFO | epoch 11 ( 70%) loss: 41948.830817\n",
      "INFO | epoch 11 (100%) loss: 40585.471777\n",
      "INFO | epoch 11 (test) loss: 39380.776062\n",
      "INFO | epoch 12 (  0%) loss: 31757.435547\n",
      "INFO | epoch 12 ( 35%) loss: 35890.404413\n",
      "INFO | epoch 12 ( 70%) loss: 34862.699403\n",
      "INFO | epoch 12 (100%) loss: 34349.373686\n",
      "INFO | epoch 12 (test) loss: 34013.087830\n",
      "INFO | epoch 13 (  0%) loss: 35969.726562\n",
      "INFO | epoch 13 ( 35%) loss: 30245.882658\n",
      "INFO | epoch 13 ( 70%) loss: 30125.416327\n",
      "INFO | epoch 13 (100%) loss: 29511.098732\n",
      "INFO | epoch 13 (test) loss: 30498.078156\n",
      "INFO | epoch 14 (  0%) loss: 25981.117188\n",
      "INFO | epoch 14 ( 35%) loss: 26263.654229\n",
      "INFO | epoch 14 ( 70%) loss: 25901.745205\n",
      "INFO | epoch 14 (100%) loss: 25409.881988\n",
      "INFO | epoch 14 (test) loss: 26366.042664\n",
      "INFO | epoch 15 (  0%) loss: 14329.519531\n",
      "INFO | epoch 15 ( 35%) loss: 22202.950978\n",
      "INFO | epoch 15 ( 70%) loss: 21480.402917\n",
      "INFO | epoch 15 (100%) loss: 21346.637199\n",
      "INFO | epoch 15 (test) loss: 22765.272034\n",
      "INFO | epoch 16 (  0%) loss: 19711.224609\n",
      "INFO | epoch 16 ( 35%) loss: 19133.793394\n",
      "INFO | epoch 16 ( 70%) loss: 18613.354667\n",
      "INFO | epoch 16 (100%) loss: 18325.774560\n",
      "INFO | epoch 16 (test) loss: 20237.426720\n",
      "INFO | epoch 17 (  0%) loss: 17046.861328\n",
      "INFO | epoch 17 ( 35%) loss: 16593.416567\n",
      "INFO | epoch 17 ( 70%) loss: 16376.621186\n",
      "INFO | epoch 17 (100%) loss: 16179.505252\n",
      "INFO | epoch 17 (test) loss: 18097.037140\n",
      "INFO | epoch 18 (  0%) loss: 14402.811523\n",
      "INFO | epoch 18 ( 35%) loss: 14671.158764\n",
      "INFO | epoch 18 ( 70%) loss: 14729.998027\n",
      "INFO | epoch 18 (100%) loss: 14566.550137\n",
      "INFO | epoch 18 (test) loss: 16820.347519\n",
      "INFO | epoch 19 (  0%) loss: 12650.464844\n",
      "INFO | epoch 19 ( 35%) loss: 12795.463727\n",
      "INFO | epoch 19 ( 70%) loss: 12925.103270\n",
      "INFO | epoch 19 (100%) loss: 12824.343866\n",
      "INFO | epoch 19 (test) loss: 15820.624786\n",
      "INFO | epoch 20 (  0%) loss: 15047.367188\n",
      "INFO | epoch 20 ( 35%) loss: 12198.136085\n",
      "INFO | epoch 20 ( 70%) loss: 12069.475103\n",
      "INFO | epoch 20 (100%) loss: 11891.168376\n",
      "INFO | epoch 20 (test) loss: 13831.914291\n",
      "INFO | epoch 21 (  0%) loss: 9320.948242\n",
      "INFO | epoch 21 ( 35%) loss: 10597.472912\n",
      "INFO | epoch 21 ( 70%) loss: 10624.656046\n",
      "INFO | epoch 21 (100%) loss: 10657.781773\n",
      "INFO | epoch 21 (test) loss: 14034.195290\n",
      "INFO | epoch 22 (  0%) loss: 11330.800781\n",
      "INFO | epoch 22 ( 35%) loss: 9356.249507\n",
      "INFO | epoch 22 ( 70%) loss: 9734.513247\n",
      "INFO | epoch 22 (100%) loss: 9780.818560\n",
      "INFO | epoch 22 (test) loss: 12447.192009\n",
      "INFO | epoch 23 (  0%) loss: 9523.190430\n",
      "INFO | epoch 23 ( 35%) loss: 9176.858921\n",
      "INFO | epoch 23 ( 70%) loss: 9161.920651\n",
      "INFO | epoch 23 (100%) loss: 9168.901567\n",
      "INFO | epoch 23 (test) loss: 11853.790611\n",
      "INFO | epoch 24 (  0%) loss: 14266.535156\n",
      "INFO | epoch 24 ( 35%) loss: 8211.944524\n",
      "INFO | epoch 24 ( 70%) loss: 8548.830197\n",
      "INFO | epoch 24 (100%) loss: 8541.055268\n",
      "INFO | epoch 24 (test) loss: 11340.909348\n",
      "INFO | epoch 25 (  0%) loss: 8817.621094\n",
      "INFO | epoch 25 ( 35%) loss: 8074.738334\n",
      "INFO | epoch 25 ( 70%) loss: 7964.035938\n",
      "INFO | epoch 25 (100%) loss: 7838.481671\n",
      "INFO | epoch 25 (test) loss: 10642.651291\n",
      "INFO | epoch 26 (  0%) loss: 8835.748047\n",
      "INFO | epoch 26 ( 35%) loss: 7664.588500\n",
      "INFO | epoch 26 ( 70%) loss: 7566.419213\n",
      "INFO | epoch 26 (100%) loss: 7594.092466\n",
      "INFO | epoch 26 (test) loss: 10174.404495\n",
      "INFO | epoch 27 (  0%) loss: 12135.719727\n",
      "INFO | epoch 27 ( 35%) loss: 6777.109162\n",
      "INFO | epoch 27 ( 70%) loss: 6897.366629\n",
      "INFO | epoch 27 (100%) loss: 6886.000787\n",
      "INFO | epoch 27 (test) loss: 9541.833923\n",
      "INFO | epoch 28 (  0%) loss: 5102.449219\n",
      "INFO | epoch 28 ( 35%) loss: 6169.500556\n",
      "INFO | epoch 28 ( 70%) loss: 6440.289461\n",
      "INFO | epoch 28 (100%) loss: 6438.840332\n",
      "INFO | epoch 28 (test) loss: 8737.636982\n",
      "INFO | epoch 29 (  0%) loss: 5272.020996\n",
      "INFO | epoch 29 ( 35%) loss: 5488.892104\n",
      "INFO | epoch 29 ( 70%) loss: 5622.726897\n",
      "INFO | epoch 29 (100%) loss: 5704.954884\n",
      "INFO | epoch 29 (test) loss: 8797.704163\n",
      "INFO | epoch 30 (  0%) loss: 6964.982422\n",
      "INFO | epoch 30 ( 35%) loss: 5536.222016\n",
      "INFO | epoch 30 ( 70%) loss: 5374.717930\n",
      "INFO | epoch 30 (100%) loss: 5281.880236\n",
      "INFO | epoch 30 (test) loss: 7641.444672\n",
      "INFO | epoch 31 (  0%) loss: 3327.297363\n",
      "INFO | epoch 31 ( 35%) loss: 4737.383455\n",
      "INFO | epoch 31 ( 70%) loss: 4639.143566\n",
      "INFO | epoch 31 (100%) loss: 4816.768736\n",
      "INFO | epoch 31 (test) loss: 7696.096493\n",
      "INFO | epoch 32 (  0%) loss: 7461.989746\n",
      "INFO | epoch 32 ( 35%) loss: 4647.351766\n",
      "INFO | epoch 32 ( 70%) loss: 4543.626577\n",
      "INFO | epoch 32 (100%) loss: 4504.436775\n",
      "INFO | epoch 32 (test) loss: 7592.783463\n",
      "INFO | epoch 33 (  0%) loss: 3001.240234\n",
      "INFO | epoch 33 ( 35%) loss: 3838.577617\n",
      "INFO | epoch 33 ( 70%) loss: 4165.557060\n",
      "INFO | epoch 33 (100%) loss: 4090.033415\n",
      "INFO | epoch 33 (test) loss: 6817.684464\n",
      "INFO | epoch 34 (  0%) loss: 2827.504883\n",
      "INFO | epoch 34 ( 35%) loss: 3745.370492\n",
      "INFO | epoch 34 ( 70%) loss: 3764.907920\n",
      "INFO | epoch 34 (100%) loss: 3731.526452\n",
      "INFO | epoch 34 (test) loss: 6534.746712\n",
      "INFO | epoch 35 (  0%) loss: 3442.815674\n",
      "INFO | epoch 35 ( 35%) loss: 3712.361592\n",
      "INFO | epoch 35 ( 70%) loss: 3581.147163\n",
      "INFO | epoch 35 (100%) loss: 3526.080395\n",
      "INFO | epoch 35 (test) loss: 6030.493908\n",
      "INFO | epoch 36 (  0%) loss: 2175.367920\n",
      "INFO | epoch 36 ( 35%) loss: 3113.270612\n",
      "INFO | epoch 36 ( 70%) loss: 3196.127588\n",
      "INFO | epoch 36 (100%) loss: 3159.422188\n",
      "INFO | epoch 36 (test) loss: 5466.576624\n",
      "INFO | epoch 37 (  0%) loss: 3895.017090\n",
      "INFO | epoch 37 ( 35%) loss: 2513.588038\n",
      "INFO | epoch 37 ( 70%) loss: 2678.900097\n",
      "INFO | epoch 37 (100%) loss: 2670.247517\n",
      "INFO | epoch 37 (test) loss: 5616.664163\n",
      "INFO | epoch 38 (  0%) loss: 2089.223145\n",
      "INFO | epoch 38 ( 35%) loss: 2646.481850\n",
      "INFO | epoch 38 ( 70%) loss: 2672.466561\n",
      "INFO | epoch 38 (100%) loss: 2679.870808\n",
      "INFO | epoch 38 (test) loss: 5688.885891\n",
      "INFO | epoch 39 (  0%) loss: 3121.147949\n",
      "INFO | epoch 39 ( 35%) loss: 2231.337617\n",
      "INFO | epoch 39 ( 70%) loss: 2370.590109\n",
      "INFO | epoch 39 (100%) loss: 2374.340127\n",
      "INFO | epoch 39 (test) loss: 5290.608746\n",
      "INFO | epoch 40 (  0%) loss: 2179.282471\n",
      "INFO | epoch 40 ( 35%) loss: 2230.397471\n",
      "INFO | epoch 40 ( 70%) loss: 2096.585034\n",
      "INFO | epoch 40 (100%) loss: 2123.483584\n",
      "INFO | epoch 40 (test) loss: 4888.415658\n",
      "INFO | epoch 41 (  0%) loss: 4512.045898\n",
      "INFO | epoch 41 ( 35%) loss: 2369.058236\n",
      "INFO | epoch 41 ( 70%) loss: 2603.096082\n",
      "INFO | epoch 41 (100%) loss: 2453.134118\n",
      "INFO | epoch 41 (test) loss: 4874.843481\n",
      "INFO | epoch 42 (  0%) loss: 2740.244385\n",
      "INFO | epoch 42 ( 35%) loss: 1781.069280\n",
      "INFO | epoch 42 ( 70%) loss: 1782.728160\n",
      "INFO | epoch 42 (100%) loss: 1818.305130\n",
      "INFO | epoch 42 (test) loss: 4540.002776\n",
      "INFO | epoch 43 (  0%) loss: 1429.253540\n",
      "INFO | epoch 43 ( 35%) loss: 1537.117650\n",
      "INFO | epoch 43 ( 70%) loss: 1562.452372\n",
      "INFO | epoch 43 (100%) loss: 1544.584495\n",
      "INFO | epoch 43 (test) loss: 4483.222338\n",
      "INFO | epoch 44 (  0%) loss: 1851.522461\n",
      "INFO | epoch 44 ( 35%) loss: 1823.921159\n",
      "INFO | epoch 44 ( 70%) loss: 1804.403606\n",
      "INFO | epoch 44 (100%) loss: 1751.196610\n",
      "INFO | epoch 44 (test) loss: 4147.958734\n",
      "INFO | epoch 45 (  0%) loss: 989.158997\n",
      "INFO | epoch 45 ( 35%) loss: 1334.770486\n",
      "INFO | epoch 45 ( 70%) loss: 1461.465630\n",
      "INFO | epoch 45 (100%) loss: 1507.118347\n",
      "INFO | epoch 45 (test) loss: 4469.360466\n",
      "INFO | epoch 46 (  0%) loss: 1081.931274\n",
      "INFO | epoch 46 ( 35%) loss: 1429.873541\n",
      "INFO | epoch 46 ( 70%) loss: 1423.329481\n",
      "INFO | epoch 46 (100%) loss: 1548.374722\n",
      "INFO | epoch 46 (test) loss: 4596.934202\n",
      "INFO | epoch 47 (  0%) loss: 1253.575928\n",
      "INFO | epoch 47 ( 35%) loss: 1861.777691\n",
      "INFO | epoch 47 ( 70%) loss: 1619.071893\n",
      "INFO | epoch 47 (100%) loss: 1511.725602\n",
      "INFO | epoch 47 (test) loss: 4227.367608\n",
      "INFO | epoch 48 (  0%) loss: 1746.462158\n",
      "INFO | epoch 48 ( 35%) loss: 1409.838460\n",
      "INFO | epoch 48 ( 70%) loss: 1340.161647\n",
      "INFO | epoch 48 (100%) loss: 1314.902592\n",
      "INFO | epoch 48 (test) loss: 3681.842785\n",
      "INFO | epoch 49 (  0%) loss: 833.600952\n",
      "INFO | epoch 49 ( 35%) loss: 1318.961595\n",
      "INFO | epoch 49 ( 70%) loss: 1293.562146\n",
      "INFO | epoch 49 (100%) loss: 1332.987604\n",
      "INFO | epoch 49 (test) loss: 4182.926363\n",
      "INFO | epoch 50 (  0%) loss: 843.650635\n",
      "INFO | epoch 50 ( 35%) loss: 1251.577425\n",
      "INFO | epoch 50 ( 70%) loss: 1193.908294\n",
      "INFO | epoch 50 (100%) loss: 1220.428266\n",
      "INFO | epoch 50 (test) loss: 3817.638424\n",
      "INFO | epoch 51 (  0%) loss: 1095.727417\n",
      "INFO | epoch 51 ( 35%) loss: 1107.240330\n",
      "INFO | epoch 51 ( 70%) loss: 1173.032806\n",
      "INFO | epoch 51 (100%) loss: 1153.131797\n",
      "INFO | epoch 51 (test) loss: 3646.418591\n",
      "INFO | epoch 52 (  0%) loss: 675.770508\n",
      "INFO | epoch 52 ( 35%) loss: 953.122190\n",
      "INFO | epoch 52 ( 70%) loss: 987.016489\n",
      "INFO | epoch 52 (100%) loss: 1018.715709\n",
      "INFO | epoch 52 (test) loss: 3449.932512\n",
      "INFO | epoch 53 (  0%) loss: 830.142578\n",
      "INFO | epoch 53 ( 35%) loss: 965.385661\n",
      "INFO | epoch 53 ( 70%) loss: 886.649986\n",
      "INFO | epoch 53 (100%) loss: 860.133365\n",
      "INFO | epoch 53 (test) loss: 3079.475058\n",
      "INFO | epoch 54 (  0%) loss: 575.500122\n",
      "INFO | epoch 54 ( 35%) loss: 757.028952\n",
      "INFO | epoch 54 ( 70%) loss: 787.797027\n",
      "INFO | epoch 54 (100%) loss: 821.497481\n",
      "INFO | epoch 54 (test) loss: 4307.336385\n",
      "INFO | epoch 55 (  0%) loss: 918.898499\n",
      "INFO | epoch 55 ( 35%) loss: 1393.123439\n",
      "INFO | epoch 55 ( 70%) loss: 1415.674322\n",
      "INFO | epoch 55 (100%) loss: 1414.935534\n",
      "INFO | epoch 55 (test) loss: 3968.709595\n",
      "INFO | epoch 56 (  0%) loss: 1385.589600\n",
      "INFO | epoch 56 ( 35%) loss: 1270.061408\n",
      "INFO | epoch 56 ( 70%) loss: 1235.506615\n",
      "INFO | epoch 56 (100%) loss: 1168.893156\n",
      "INFO | epoch 56 (test) loss: 3795.003469\n",
      "INFO | epoch 57 (  0%) loss: 1425.763306\n",
      "INFO | epoch 57 ( 35%) loss: 816.702482\n",
      "INFO | epoch 57 ( 70%) loss: 835.323728\n",
      "INFO | epoch 57 (100%) loss: 830.392921\n",
      "INFO | epoch 57 (test) loss: 3434.232013\n",
      "INFO | epoch 58 (  0%) loss: 993.592041\n",
      "INFO | epoch 58 ( 35%) loss: 812.127251\n",
      "INFO | epoch 58 ( 70%) loss: 845.236124\n",
      "INFO | epoch 58 (100%) loss: 833.751578\n",
      "INFO | epoch 58 (test) loss: 3194.967657\n",
      "INFO | epoch 59 (  0%) loss: 640.245972\n",
      "INFO | epoch 59 ( 35%) loss: 629.989435\n",
      "INFO | epoch 59 ( 70%) loss: 670.792152\n",
      "INFO | epoch 59 (100%) loss: 720.041718\n",
      "INFO | epoch 59 (test) loss: 4309.616161\n",
      "INFO | epoch 60 (  0%) loss: 3316.812744\n",
      "INFO | epoch 60 ( 35%) loss: 1129.538542\n",
      "INFO | epoch 60 ( 70%) loss: 1012.120964\n",
      "INFO | epoch 60 (100%) loss: 1083.745498\n",
      "INFO | epoch 60 (test) loss: 3769.038116\n",
      "INFO | epoch 61 (  0%) loss: 1039.730225\n",
      "INFO | epoch 61 ( 35%) loss: 820.382921\n",
      "INFO | epoch 61 ( 70%) loss: 876.383603\n",
      "INFO | epoch 61 (100%) loss: 908.358465\n",
      "INFO | epoch 61 (test) loss: 3500.315755\n",
      "INFO | epoch 62 (  0%) loss: 687.243042\n",
      "INFO | epoch 62 ( 35%) loss: 883.049425\n",
      "INFO | epoch 62 ( 70%) loss: 872.376586\n",
      "INFO | epoch 62 (100%) loss: 858.710030\n",
      "INFO | epoch 62 (test) loss: 2864.292210\n",
      "INFO | epoch 63 (  0%) loss: 588.840820\n",
      "INFO | epoch 63 ( 35%) loss: 729.161786\n",
      "INFO | epoch 63 ( 70%) loss: 749.356507\n",
      "INFO | epoch 63 (100%) loss: 747.174512\n",
      "INFO | epoch 63 (test) loss: 3086.375848\n",
      "INFO | epoch 64 (  0%) loss: 706.356628\n",
      "INFO | epoch 64 ( 35%) loss: 573.587830\n",
      "INFO | epoch 64 ( 70%) loss: 615.156228\n",
      "INFO | epoch 64 (100%) loss: 634.530748\n",
      "INFO | epoch 64 (test) loss: 2930.764375\n",
      "INFO | epoch 65 (  0%) loss: 354.810883\n",
      "INFO | epoch 65 ( 35%) loss: 826.870498\n",
      "INFO | epoch 65 ( 70%) loss: 742.859040\n",
      "INFO | epoch 65 (100%) loss: 685.786961\n",
      "INFO | epoch 65 (test) loss: 2610.220900\n",
      "INFO | epoch 66 (  0%) loss: 580.327332\n",
      "INFO | epoch 66 ( 35%) loss: 494.813944\n",
      "INFO | epoch 66 ( 70%) loss: 489.477089\n",
      "INFO | epoch 66 (100%) loss: 475.017625\n",
      "INFO | epoch 66 (test) loss: 2778.848577\n",
      "INFO | epoch 67 (  0%) loss: 453.841492\n",
      "INFO | epoch 67 ( 35%) loss: 449.515378\n",
      "INFO | epoch 67 ( 70%) loss: 491.562671\n",
      "INFO | epoch 67 (100%) loss: 481.343048\n",
      "INFO | epoch 67 (test) loss: 3007.432701\n",
      "INFO | epoch 68 (  0%) loss: 288.504395\n",
      "INFO | epoch 68 ( 35%) loss: 464.958453\n",
      "INFO | epoch 68 ( 70%) loss: 532.686417\n",
      "INFO | epoch 68 (100%) loss: 516.161379\n",
      "INFO | epoch 68 (test) loss: 2709.216156\n",
      "INFO | epoch 69 (  0%) loss: 685.523132\n",
      "INFO | epoch 69 ( 35%) loss: 856.671895\n",
      "INFO | epoch 69 ( 70%) loss: 834.024777\n",
      "INFO | epoch 69 (100%) loss: 951.834012\n",
      "INFO | epoch 69 (test) loss: 5023.451353\n",
      "INFO | epoch 70 (  0%) loss: 1610.327881\n",
      "INFO | epoch 70 ( 35%) loss: 1857.683645\n",
      "INFO | epoch 70 ( 70%) loss: 1592.683598\n",
      "INFO | epoch 70 (100%) loss: 1378.807274\n",
      "INFO | epoch 70 (test) loss: 3377.654554\n",
      "INFO | epoch 71 (  0%) loss: 416.198059\n",
      "INFO | epoch 71 ( 35%) loss: 696.934226\n",
      "INFO | epoch 71 ( 70%) loss: 633.557479\n",
      "INFO | epoch 71 (100%) loss: 608.643068\n",
      "INFO | epoch 71 (test) loss: 3032.799382\n",
      "INFO | epoch 72 (  0%) loss: 353.729126\n",
      "INFO | epoch 72 ( 35%) loss: 449.819456\n",
      "INFO | epoch 72 ( 70%) loss: 477.861616\n",
      "INFO | epoch 72 (100%) loss: 470.032907\n",
      "INFO | epoch 72 (test) loss: 2872.961200\n",
      "INFO | epoch 73 (  0%) loss: 367.421417\n",
      "INFO | epoch 73 ( 35%) loss: 369.725620\n",
      "INFO | epoch 73 ( 70%) loss: 358.939084\n",
      "INFO | epoch 73 (100%) loss: 344.667009\n",
      "INFO | epoch 73 (test) loss: 2307.105837\n",
      "INFO | epoch 74 (  0%) loss: 270.305145\n",
      "INFO | epoch 74 ( 35%) loss: 284.496873\n",
      "INFO | epoch 74 ( 70%) loss: 283.003658\n",
      "INFO | epoch 74 (100%) loss: 280.625616\n",
      "INFO | epoch 74 (test) loss: 2396.454479\n",
      "INFO | epoch 75 (  0%) loss: 279.022430\n",
      "INFO | epoch 75 ( 35%) loss: 264.134211\n",
      "INFO | epoch 75 ( 70%) loss: 258.345012\n",
      "INFO | epoch 75 (100%) loss: 253.124255\n",
      "INFO | epoch 75 (test) loss: 2410.402098\n",
      "INFO | epoch 76 (  0%) loss: 219.211945\n",
      "INFO | epoch 76 ( 35%) loss: 241.448362\n",
      "INFO | epoch 76 ( 70%) loss: 240.270678\n",
      "INFO | epoch 76 (100%) loss: 273.773707\n",
      "INFO | epoch 76 (test) loss: 2750.511723\n",
      "INFO | epoch 77 (  0%) loss: 276.347900\n",
      "INFO | epoch 77 ( 35%) loss: 396.937387\n",
      "INFO | epoch 77 ( 70%) loss: 563.016945\n",
      "INFO | epoch 77 (100%) loss: 648.436571\n",
      "INFO | epoch 77 (test) loss: 3689.515570\n",
      "INFO | epoch 78 (  0%) loss: 509.605499\n",
      "INFO | epoch 78 ( 35%) loss: 1525.683228\n",
      "INFO | epoch 78 ( 70%) loss: 1234.839111\n",
      "INFO | epoch 78 (100%) loss: 1111.536049\n",
      "INFO | epoch 78 (test) loss: 3691.108822\n",
      "INFO | epoch 79 (  0%) loss: 696.555664\n",
      "INFO | epoch 79 ( 35%) loss: 630.233048\n",
      "INFO | epoch 79 ( 70%) loss: 567.554170\n",
      "INFO | epoch 79 (100%) loss: 527.187777\n",
      "INFO | epoch 79 (test) loss: 2658.135778\n",
      "INFO | epoch 80 (  0%) loss: 368.542847\n",
      "INFO | epoch 80 ( 35%) loss: 343.248234\n",
      "INFO | epoch 80 ( 70%) loss: 325.822841\n",
      "INFO | epoch 80 (100%) loss: 327.829780\n",
      "INFO | epoch 80 (test) loss: 2524.338593\n",
      "INFO | epoch 81 (  0%) loss: 239.805237\n",
      "INFO | epoch 81 ( 35%) loss: 340.460822\n",
      "INFO | epoch 81 ( 70%) loss: 307.698686\n",
      "INFO | epoch 81 (100%) loss: 298.395139\n",
      "INFO | epoch 81 (test) loss: 2501.184160\n",
      "INFO | epoch 82 (  0%) loss: 327.608765\n",
      "INFO | epoch 82 ( 35%) loss: 341.364440\n",
      "INFO | epoch 82 ( 70%) loss: 313.184676\n",
      "INFO | epoch 82 (100%) loss: 289.206871\n",
      "INFO | epoch 82 (test) loss: 2366.930183\n",
      "INFO | epoch 83 (  0%) loss: 340.454651\n",
      "INFO | epoch 83 ( 35%) loss: 299.354985\n",
      "INFO | epoch 83 ( 70%) loss: 322.162828\n",
      "INFO | epoch 83 (100%) loss: 313.188009\n",
      "INFO | epoch 83 (test) loss: 2317.451898\n",
      "INFO | epoch 84 (  0%) loss: 220.288101\n",
      "INFO | epoch 84 ( 35%) loss: 250.370629\n",
      "INFO | epoch 84 ( 70%) loss: 306.629858\n",
      "INFO | epoch 84 (100%) loss: 311.388975\n",
      "INFO | epoch 84 (test) loss: 2836.671191\n",
      "INFO | epoch 85 (  0%) loss: 264.217834\n",
      "INFO | epoch 85 ( 35%) loss: 509.980990\n",
      "INFO | epoch 85 ( 70%) loss: 740.283558\n",
      "INFO | epoch 85 (100%) loss: 757.738074\n",
      "INFO | epoch 85 (test) loss: 4287.348223\n",
      "INFO | epoch 86 (  0%) loss: 1618.635132\n",
      "INFO | epoch 86 ( 35%) loss: 754.633818\n",
      "INFO | epoch 86 ( 70%) loss: 741.324094\n",
      "INFO | epoch 86 (100%) loss: 712.014071\n",
      "INFO | epoch 86 (test) loss: 2581.449963\n",
      "INFO | epoch 87 (  0%) loss: 290.736145\n",
      "INFO | epoch 87 ( 35%) loss: 459.788459\n",
      "INFO | epoch 87 ( 70%) loss: 428.341030\n",
      "INFO | epoch 87 (100%) loss: 406.925430\n",
      "INFO | epoch 87 (test) loss: 2223.810423\n",
      "INFO | epoch 88 (  0%) loss: 217.617096\n",
      "INFO | epoch 88 ( 35%) loss: 245.753209\n",
      "INFO | epoch 88 ( 70%) loss: 254.614423\n",
      "INFO | epoch 88 (100%) loss: 258.055740\n",
      "INFO | epoch 88 (test) loss: 2137.898087\n",
      "INFO | epoch 89 (  0%) loss: 199.060303\n",
      "INFO | epoch 89 ( 35%) loss: 226.010874\n",
      "INFO | epoch 89 ( 70%) loss: 218.414945\n",
      "INFO | epoch 89 (100%) loss: 209.590835\n",
      "INFO | epoch 89 (test) loss: 2145.025507\n",
      "INFO | epoch 90 (  0%) loss: 231.466629\n",
      "INFO | epoch 90 ( 35%) loss: 199.061093\n",
      "INFO | epoch 90 ( 70%) loss: 267.973409\n",
      "INFO | epoch 90 (100%) loss: 329.940643\n",
      "INFO | epoch 90 (test) loss: 2376.258435\n",
      "INFO | epoch 91 (  0%) loss: 461.014832\n",
      "INFO | epoch 91 ( 35%) loss: 436.959697\n",
      "INFO | epoch 91 ( 70%) loss: 417.798056\n",
      "INFO | epoch 91 (100%) loss: 403.631984\n",
      "INFO | epoch 91 (test) loss: 2398.490468\n",
      "INFO | epoch 92 (  0%) loss: 198.699829\n",
      "INFO | epoch 92 ( 35%) loss: 535.714597\n",
      "INFO | epoch 92 ( 70%) loss: 580.560784\n",
      "INFO | epoch 92 (100%) loss: 521.677848\n",
      "INFO | epoch 92 (test) loss: 2548.618070\n",
      "INFO | epoch 93 (  0%) loss: 245.245117\n",
      "INFO | epoch 93 ( 35%) loss: 323.981696\n",
      "INFO | epoch 93 ( 70%) loss: 395.684129\n",
      "INFO | epoch 93 (100%) loss: 428.390097\n",
      "INFO | epoch 93 (test) loss: 3042.952134\n",
      "INFO | epoch 94 (  0%) loss: 204.649734\n",
      "INFO | epoch 94 ( 35%) loss: 7232425.169236\n",
      "INFO | epoch 94 ( 70%) loss: 3634766.086345\n",
      "INFO | epoch 94 (100%) loss: 2590930.334629\n",
      "INFO | epoch 94 (test) loss: 2461.701287\n",
      "INFO | epoch 95 (  0%) loss: 487.139984\n",
      "INFO | epoch 95 ( 35%) loss: 497.392672\n",
      "INFO | epoch 95 ( 70%) loss: 466.135381\n",
      "INFO | epoch 95 (100%) loss: 1204.293068\n",
      "INFO | epoch 95 (test) loss: 3059.828469\n",
      "INFO | epoch 96 (  0%) loss: 397.087708\n",
      "INFO | epoch 96 ( 35%) loss: 376.010384\n",
      "INFO | epoch 96 ( 70%) loss: 344.314221\n",
      "INFO | epoch 96 (100%) loss: 338.594827\n",
      "INFO | epoch 96 (test) loss: 2138.227397\n",
      "INFO | epoch 97 (  0%) loss: 249.804108\n",
      "INFO | epoch 97 ( 35%) loss: 273.094362\n",
      "INFO | epoch 97 ( 70%) loss: 261.302807\n",
      "INFO | epoch 97 (100%) loss: 262.448951\n",
      "INFO | epoch 97 (test) loss: 2297.935203\n",
      "INFO | epoch 98 (  0%) loss: 163.500580\n",
      "INFO | epoch 98 ( 35%) loss: 250.932314\n",
      "INFO | epoch 98 ( 70%) loss: 259.235164\n",
      "INFO | epoch 98 (100%) loss: 247.277005\n",
      "INFO | epoch 98 (test) loss: 2125.729264\n",
      "INFO | epoch 99 (  0%) loss: 181.090027\n",
      "INFO | epoch 99 ( 35%) loss: 248.700197\n",
      "INFO | epoch 99 ( 70%) loss: 221.649986\n",
      "INFO | epoch 99 (100%) loss: 216.018389\n",
      "INFO | epoch 99 (test) loss: 2213.926233\n",
      "INFO | epoch 100 (  0%) loss: 201.498795\n",
      "INFO | epoch 100 ( 35%) loss: 200.326215\n",
      "INFO | epoch 100 ( 70%) loss: 189.532617\n",
      "INFO | epoch 100 (100%) loss: 187.026668\n",
      "INFO | epoch 100 (test) loss: 2067.567595\n",
      "INFO | model blob saved to /home/jmsdnns/ML/pallets/saved/cvae.conv.labeled.pkl\n",
      "INFO | training losses saved to /home/jmsdnns/ML/pallets/saved/cvae.conv.labeled.json\n"
     ]
    }
   ],
   "source": [
    "model = M.cvae.LabeledConvCVAE(input_dim, latent_dim, classes_dim)\n",
    "criterion = M.cvae.Loss()\n",
    "\n",
    "train_losses, test_losses = M.cvae.train(\n",
    "    device, model, criterion, train_loader, test_loader,\n",
    "    learn_rate=LR, epochs=EPOCHS, conditional_loss=True\n",
    ")\n",
    "\n",
    "M.save(SAVE_NAME, model, train_losses, test_losses)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77fb32ad",
   "metadata": {},
   "source": [
    "# Results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18005438",
   "metadata": {},
   "source": [
    "## Random Inputs w/ Real Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f2b9d212",
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "\n",
    "def rand_label():\n",
    "    label_idx = int(random.random() * len(dataset._labels))\n",
    "    features = dataset._labels[label_idx]\n",
    "    enabled_names = [k for k,v in zip(dataset._label_keys, features) if v.item() == 1]\n",
    "    return features.to(device), enabled_names\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "38c47d81",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generate new image\n",
    "\n",
    "def rand_punk():\n",
    "    z = torch.randn(1, latent_dim).to(device)\n",
    "    # print(z.shape)\n",
    "    features, names = rand_label()\n",
    "    print(f\"Features: {', '.join(names)}\")\n",
    "\n",
    "    with torch.no_grad():\n",
    "        model.eval()\n",
    "        generated_image = model.decoder(z, features.unsqueeze(0))\n",
    "        # print(generated_image.shape)\n",
    "\n",
    "    decoded_one_hot = generated_image[0][:222, :, :]\n",
    "    decoded_one_hot = decoded_one_hot.view((222, 24, 24))\n",
    "    # print(decoded_one_hot.shape)\n",
    "    decoded = DS.one_hot_to_rgba(decoded_one_hot, dataset.mapper)\n",
    "    # print(decoded.shape)\n",
    "    return decoded\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31e74ceb",
   "metadata": {},
   "source": [
    "### 5 Randoms w/ Real Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: male, luxuriousBeard, frumpyHair, bigShades\n",
      "Shape: torch.Size([4, 24, 24])\n",
      "torch.Size([4, 24, 24])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded = rand_punk()\n",
    "print(decoded.shape)\n",
    "\n",
    "plt.imshow(transforms.functional.to_pil_image(decoded))\n",
    "plt.axis('off')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6744444a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: female, pilotHelmet, cigarette\n",
      "Shape: torch.Size([4, 24, 24])\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAGFCAYAAAASI+9IAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAGtUlEQVR4nO3doW5lRRzA4R7SJ7hYxBoIhhoeARL8PgGlkto1hIRdvW7Tyu4GxaomPAC2DtM1BAwBXyTyYMgvNZtAz9nO7b3f5yfzz0l7fxkzM83zPB8AwMHBwXujBwBge4gCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFADI4egB2D1HpyejR9gr12cXo0dghzgpABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAMs3zPI8egu0yTdPoEbhHfgK4zUkBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgh6MH4N1Y8ibC099+XXGS+/X0w4/G7PuAv9mSvxVvMeweJwUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBcnc1WGXX19VJL537IV2+zW5wUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAMs3zPI8egvVN0zR6hCE++fqrIfu+OX85ZN/R/HzsHicFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBADkcPwLsx6grpfeV737+bm5s7r91sNitOslucFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFAOI9hS11dHoybO9nj+Zhe3P/vvt9Gj0CW8RJAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFdns7pffv5p2N4ff/rZsL2XGPnNDt7/fNzebB0nBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQFydzeq+ef3Hndf+8O0XK07ycAz9Zn8tW85ucVIAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACI9xRY3eXz49EjPDhDv5n3FLjFSQGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBXZ2+p67OLReuPTk9WmgTYJ04KAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIA8Z4C/Ovxk1d3Xnv5/HjFSfgvNpvN6BF2kpMCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAg0zzP8+ghWN/R6cmd1z575E/ivi25tpv/z8/e2zkpABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAcjh6AFjTjy9eD9n3+z//XrT+6upqpUn2xwfnX9557TRNi/be5au3nRQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAyzbt8MfgeOzo9ufPaN+cvF+19+fx40fp99PjJq9Ej7BU/e2/npABARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAMjh6AHgtqVXSO/jtd1Lr4Fecs369dnFor3ZPk4KAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAmeall7Gzc6ZpWrR+5JsGS99jGMW/IdvCSQGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoAJDD0QPAmpZc273k2m1XX7MrnBQAiCgAEFEAIKIAQEQBgIgCABEFACIKAEQUAIgoABBRACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAgogBARAGAiAIAEQUAIgoARBQAiCgAEFEAIKIAQEQBgIgCAJnmeZ5HDwHAdnBSACCiAEBEAYCIAgARBQAiCgBEFACIKAAQUQAg/wCCIIeWdIttSQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded = rand_punk()\n",
    "\n",
    "plt.imshow(transforms.functional.to_pil_image(decoded))\n",
    "plt.axis('off')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e190c583",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: female, stringyHair\n",
      "Shape: torch.Size([4, 24, 24])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded = rand_punk()\n",
    "\n",
    "plt.imshow(transforms.functional.to_pil_image(decoded))\n",
    "plt.axis('off')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "be357c27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: male, normalBeardBlack, eyeMask\n",
      "Shape: torch.Size([4, 24, 24])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded = rand_punk()\n",
    "\n",
    "plt.imshow(transforms.functional.to_pil_image(decoded))\n",
    "plt.axis('off')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "d8f339ae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: male, buckTeeth, messyHair, nerdGlasses\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape: torch.Size([4, 24, 24])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "decoded = rand_punk()\n",
    "\n",
    "plt.imshow(transforms.functional.to_pil_image(decoded))\n",
    "plt.axis('off')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e1bd761",
   "metadata": {},
   "source": [
    "## Reconstruction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a150efab",
   "metadata": {},
   "outputs": [],
   "source": [
    "def reconstruct_punk(idx):\n",
    "    punk = I.get_punk_tensor(idx)\n",
    "    p = DS.rgba_to_one_hot(punk, dataset.mapper)\n",
    "    p = p.unsqueeze(0)\n",
    "    p = p.to(device)\n",
    "\n",
    "    _, labels = dataset[idx]\n",
    "    l = labels.unsqueeze(0)\n",
    "    l = l.to(device)\n",
    "    enabled_features = [k for k,v in zip(dataset._label_keys, labels) if v.item() == 1]\n",
    "\n",
    "    model.eval()\n",
    "    with torch.no_grad():\n",
    "        reconstructed, mu, logvar = model.forward(p, l)\n",
    "\n",
    "    recon_punk = reconstructed[0][:222, :, :].cpu()\n",
    "    recon_punk = DS.one_hot_to_rgba(recon_punk, dataset.mapper)\n",
    "\n",
    "    return punk, recon_punk, enabled_features\n",
    "\n",
    "\n",
    "def draw_two(img1, img2):\n",
    "    page_size = 2\n",
    "    view_x, view_y = 4*page_size, 2*page_size\n",
    "    fig = plt.figure(figsize=(view_x, view_y))\n",
    "    fig.add_subplot(1, 2, 1)\n",
    "    plt.imshow(transforms.functional.to_pil_image(img1))\n",
    "    plt.axis('off')\n",
    "    fig.add_subplot(1, 2, 2)\n",
    "    plt.imshow(transforms.functional.to_pil_image(img2))\n",
    "    plt.axis('off')\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a12b220",
   "metadata": {},
   "source": [
    "### 5 Reconstructions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "142e642d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: female, rosyCheeks, darkHair, cigarette\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "punk, recon_punk, features = reconstruct_punk(1000)\n",
    "\n",
    "print(f\"Features: {', '.join(features)}\")\n",
    "draw_two(punk, recon_punk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: female, stringyHair, purpleLipstick, earring\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "punk, recon_punk, features = reconstruct_punk(2001)\n",
    "\n",
    "print(f\"Features: {', '.join(features)}\")\n",
    "draw_two(punk, recon_punk)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8e401eca",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: female, wildWhiteHair, greenEyeShadow, hotLipstick\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAn8AAAEtCAYAAAB54AaaAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAG40lEQVR4nO3dsW0UURRA0RnkHhA5zsgoggjcBgmIDrx0gIgogAYMEX2QmQaoYsiIMNovmJ3ZvefEX6O/wT5d/eTNy7IsEwAACY+2vgAAAKcj/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCrra+ANv58eX90Pmnr25XugnAeTJHOUde/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPVuF2Zk1dDomqG11xiNfn+ElUrAsczRPzNHL4eXPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELs9g1bcwfk3qy9TxNoMkcfZo7ul5c/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQuz23bnRXYrvv39b6SbTdPvsxWrfHrXm75ymafp8cxg6vyzLOhcB/pk5+mfmaJeXPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELmxTK9kxrdMTnqemCX4ttPL4e+/ebJ88HbjHn66vbos/M8D3179Ld+fP116Ly/EZyOOfowc5RjePkDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHWu53Y6Dqd+7vDOhfhr0bWO02TtURwSuboeTBH98vLHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQcrX1Bbhc1zeHofP3d2PnAS6dOcoavPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQYrcvq7FjEuDfmKOswcsfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAix3o2j/Xz3Yesr/Pb4w7utrwAwzBxlD7z8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAEDIvy7JsfQkeNs/z0Pn7u8M6Fzlz1zeHofP+FnA5zNH/wxy9HF7+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAECI+AMACLna+gJsZ3RP4572XdoxCeyBOco58vIHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiN2+YaM7Jkf3QK7JjklgD8xRzpGXPwCAEPEHABAi/gAAQsQfAECI+AMACBF/AAAh4g8AIET8AQCEiD8AgBDxBwAQYr0bR1t7jZFVQ8ClM0fZAy9/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIeIPACBE/AEAhIg/AIAQ8QcAECL+AABCxB8AQIj4AwAIEX8AACHiDwAgRPwBAISIPwCAEPEHABAi/gAAQsQfAEDIvCzLsvUlAAA4DS9/AAAh4g8AIET8AQCEiD8AgBDxBwAQIv4AAELEHwBAiPgDAAgRfwAAIb8AMqy1fDjPnecAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "punk, recon_punk, features = reconstruct_punk(5000)\n",
    "\n",
    "print(f\"Features: {', '.join(features)}\")\n",
    "draw_two(punk, recon_punk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "99f3ac30",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: male, rosyCheeks, mustache, vr, cap, cigarette\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "punk, recon_punk, features = reconstruct_punk(8000)\n",
    "\n",
    "print(f\"Features: {', '.join(features)}\")\n",
    "draw_two(punk, recon_punk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "eb402159",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features: male, handlebars, bandana\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "punk, recon_punk, features = reconstruct_punk(1337)\n",
    "\n",
    "print(f\"Features: {', '.join(features)}\")\n",
    "draw_two(punk, recon_punk)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04a97510",
   "metadata": {},
   "source": [
    "## Losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show(close=None, block=None)>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title(\"Train & Test loss\")\n",
    "plt.plot(train_losses, label='train loss')\n",
    "plt.plot(test_losses, label='test loss')\n",
    "plt.legend()\n",
    "plt.show\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "c1261252",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[425156.3146054965,\n",
       " 267729.86642564274,\n",
       " 231887.83935893173,\n",
       " 213834.91061336437,\n",
       " 206240.89983654697,\n",
       " 192189.33597905585,\n",
       " 173946.75849124556,\n",
       " 150144.55444509088,\n",
       " 122305.58777288342,\n",
       " 98320.38017370345,\n",
       " 81170.9435533023,\n",
       " 68698.74737159241,\n",
       " 59022.19746301529,\n",
       " 50819.763976617905,\n",
       " 42693.27439847906,\n",
       " 36651.54911901596,\n",
       " 32359.010503241356,\n",
       " 29133.100274268618,\n",
       " 25648.687732020167,\n",
       " 23782.33675216783,\n",
       " 21315.563546687998,\n",
       " 19561.63712045656,\n",
       " 18337.803134350066,\n",
       " 17082.110536832337,\n",
       " 15676.963342544881,\n",
       " 15188.184932194703,\n",
       " 13772.001573927859,\n",
       " 12877.6806640625,\n",
       " 11409.90976839539,\n",
       " 10563.760472940216,\n",
       " 9633.537471257203,\n",
       " 9008.873549873948,\n",
       " 8180.066829167359,\n",
       " 7463.052904927139,\n",
       " 7052.160789976729,\n",
       " 6318.84437571857,\n",
       " 5340.495033210051,\n",
       " 5359.741616972795,\n",
       " 4748.680254564218,\n",
       " 4246.967168713292,\n",
       " 4906.268235659769,\n",
       " 3636.610259793329,\n",
       " 3089.1689903313386,\n",
       " 3502.393220617416,\n",
       " 3014.236694552374,\n",
       " 3096.7494442987104,\n",
       " 3023.4512034747618,\n",
       " 2629.805183897627,\n",
       " 2665.975208904726,\n",
       " 2440.856532752937,\n",
       " 2306.263594823526,\n",
       " 2037.4314170458638,\n",
       " 1720.266730342351,\n",
       " 1642.9949613530584,\n",
       " 2829.8710675611565,\n",
       " 2337.786311805671,\n",
       " 1660.785841164014,\n",
       " 1667.5031562967504,\n",
       " 1440.0834357079038,\n",
       " 2167.490996448706,\n",
       " 1816.7169302000223,\n",
       " 1717.4200595287566,\n",
       " 1494.3490244006434,\n",
       " 1269.0614963829094,\n",
       " 1371.573922880998,\n",
       " 950.0352490472455,\n",
       " 962.6860961372971,\n",
       " 1032.322758938404,\n",
       " 1903.6680239413647,\n",
       " 2757.6145487034573,\n",
       " 1217.2861356261774,\n",
       " 940.0658137544673,\n",
       " 689.3340175980372,\n",
       " 561.2512321742714,\n",
       " 506.2485108612277,\n",
       " 547.5474140627165,\n",
       " 1296.8731422153771,\n",
       " 2223.0720974536653,\n",
       " 1054.375554943761,\n",
       " 655.6595600750429,\n",
       " 596.79027857679,\n",
       " 578.4137425524124,\n",
       " 626.3760180101327,\n",
       " 622.7779504221381,\n",
       " 1515.476148672983,\n",
       " 1424.0281414275473,\n",
       " 813.8508592970828,\n",
       " 516.111480361181,\n",
       " 419.18167073675926,\n",
       " 659.8812851060367,\n",
       " 807.2639675275653,\n",
       " 1043.3556966578706,\n",
       " 856.7801937887855,\n",
       " 5181860.669257523,\n",
       " 2408.5861369464415,\n",
       " 677.1896533289699,\n",
       " 524.8979012239064,\n",
       " 494.55401091879986,\n",
       " 432.03677790215676,\n",
       " 374.0533363937486]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_losses\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "15d6f86b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[270388.5743408203,\n",
       " 259954.48486328125,\n",
       " 214847.9969482422,\n",
       " 208389.3232421875,\n",
       " 200367.4149169922,\n",
       " 179334.4100341797,\n",
       " 162466.49536132812,\n",
       " 133166.53674316406,\n",
       " 110549.38653564453,\n",
       " 95741.35089111328,\n",
       " 78761.55212402344,\n",
       " 68026.17565917969,\n",
       " 60996.156311035156,\n",
       " 52732.08532714844,\n",
       " 45530.54406738281,\n",
       " 40474.853439331055,\n",
       " 36194.074279785156,\n",
       " 33640.6950378418,\n",
       " 31641.249572753906,\n",
       " 27663.828582763672,\n",
       " 28068.390579223633,\n",
       " 24894.384017944336,\n",
       " 23707.58122253418,\n",
       " 22681.81869506836,\n",
       " 21285.30258178711,\n",
       " 20348.808990478516,\n",
       " 19083.667846679688,\n",
       " 17475.273963928223,\n",
       " 17595.408325195312,\n",
       " 15282.889343261719,\n",
       " 15392.192985534668,\n",
       " 15185.566925048828,\n",
       " 13635.368927001953,\n",
       " 13069.493423461914,\n",
       " 12060.987815856934,\n",
       " 10933.153247833252,\n",
       " 11233.328325271606,\n",
       " 11377.771781921387,\n",
       " 10581.217491149902,\n",
       " 9776.831315994263,\n",
       " 9749.686962127686,\n",
       " 9080.00555229187,\n",
       " 8966.444675445557,\n",
       " 8295.91746711731,\n",
       " 8938.720932006836,\n",
       " 9193.868404388428,\n",
       " 8454.735216140747,\n",
       " 7363.685569763184,\n",
       " 8365.852725982666,\n",
       " 7635.2768478393555,\n",
       " 7292.837181091309,\n",
       " 6899.86502456665,\n",
       " 6158.950116157532,\n",
       " 8614.672769546509,\n",
       " 7937.419189453125,\n",
       " 7590.006938934326,\n",
       " 6868.4640254974365,\n",
       " 6389.935314178467,\n",
       " 8619.232322692871,\n",
       " 7538.076232910156,\n",
       " 7000.631509780884,\n",
       " 5728.584419250488,\n",
       " 6172.751695156097,\n",
       " 5861.528750896454,\n",
       " 5220.441799163818,\n",
       " 5557.697153091431,\n",
       " 6014.86540222168,\n",
       " 5418.432312011719,\n",
       " 10046.90270614624,\n",
       " 6755.309108734131,\n",
       " 6065.598764419556,\n",
       " 5745.922400474548,\n",
       " 4614.211674690247,\n",
       " 4792.908958435059,\n",
       " 4820.804195404053,\n",
       " 5501.023446559906,\n",
       " 7379.031139373779,\n",
       " 7382.217643737793,\n",
       " 5316.271556854248,\n",
       " 5048.677186012268,\n",
       " 5002.368319511414,\n",
       " 4733.860366344452,\n",
       " 4634.90379524231,\n",
       " 5673.34238243103,\n",
       " 8574.696445465088,\n",
       " 5162.899925231934,\n",
       " 4447.620846748352,\n",
       " 4275.796173810959,\n",
       " 4290.0510149002075,\n",
       " 4752.516870975494,\n",
       " 4796.980936050415,\n",
       " 5097.236140966415,\n",
       " 6085.904267549515,\n",
       " 4923.402573108673,\n",
       " 6119.656937599182,\n",
       " 4276.454794883728,\n",
       " 4595.870406389236,\n",
       " 4251.458527088165,\n",
       " 4427.852466344833,\n",
       " 4135.135190010071]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_losses\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2c85eb18",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}