custom_eval_nlayers.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/creddy/Work/vae /home/creddy/Work/vae/./data/numbers_changed/\n",
      "Using Differential Compositing\n",
      "loading:  last.ckpt\n"
     ]
    }
   ],
   "source": [
    "import yaml\n",
    "import argparse\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "from models import *\n",
    "from experiment import VAEXperiment\n",
    "import torch\n",
    "import torch.backends.cudnn as cudnn\n",
    "from pytorch_lightning import Trainer\n",
    "from pytorch_lightning.loggers import TestTubeLogger\n",
    "from pytorch_lightning.callbacks import ModelCheckpoint\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import torchvision.utils as vutils\n",
    "\n",
    "filename = './configs/numbers_soft.yaml'\n",
    "with open(filename, 'r') as file:\n",
    "    try:\n",
    "        config = yaml.safe_load(file)\n",
    "    except yaml.YAMLError as exc:\n",
    "        print(exc)\n",
    "\n",
    "# For reproducibility\n",
    "torch.manual_seed(config['logging_params']['manual_seed'])\n",
    "np.random.seed(config['logging_params']['manual_seed'])\n",
    "cudnn.deterministic = True\n",
    "cudnn.benchmark = False\n",
    "\n",
    "model_save_path = os.getcwd()#'{}/{}/version_{}'.format(config['logging_params']['save_dir'], config['logging_params']['name'], tt_logger.version)\n",
    "parent = '/'.join(model_save_path.split('/')[:-3])\n",
    "config['logging_params']['save_dir'] = os.path.join(parent, config['logging_params']['save_dir'])\n",
    "config['exp_params']['data_path'] = os.path.join(parent, config['exp_params']['data_path'])\n",
    "print(parent, config['exp_params']['data_path'])\n",
    "\n",
    "model = vae_models[config['model_params']['name']](imsize=config['exp_params']['img_size'], **config['model_params'])\n",
    "experiment = VAEXperiment(model,\n",
    "                          config['exp_params'])\n",
    "\n",
    "weights = [x for x in os.listdir(model_save_path) if '.ckpt' in x]\n",
    "weights.sort(key=lambda x: os.path.getmtime(x))\n",
    "load_weight = weights[-1]\n",
    "print('loading: ', load_weight)\n",
    "\n",
    "checkpoint = torch.load(load_weight)\n",
    "experiment.load_state_dict(checkpoint['state_dict'])\n",
    "train_dataloader = experiment.train_dataloader()\n",
    "experiment.eval()\n",
    "experiment.freeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def interpolate_mini(experiment, x, **kwargs) -> Tensor:\n",
    "    mu, log_var = experiment.model.encode(x)\n",
    "    z = experiment.model.interpolate_vectors(mu[0], mu[1], 10)\n",
    "    output = experiment.model.decode_and_composite(z, verbose=kwargs['verbose'])\n",
    "    return output\n",
    "\n",
    "\n",
    "def search_samples(search_string):\n",
    "    for idx, name in enumerate(experiment.sample_dataloader.dataset.samples):\n",
    "        if search_string in name[0]:\n",
    "            return idx\n",
    "    return 'NA'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "idx1 = search_samples('0_Akashi_6_8.png')\n",
    "idx2 = search_samples('0_Akashi_4_8.png')\n",
    "\n",
    "x1, _ = experiment.sample_dataloader.dataset[idx1]\n",
    "x2, _ = experiment.sample_dataloader.dataset[idx2]\n",
    "\n",
    "x = torch.stack([x1, x2], dim=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/creddy/anaconda3/envs/sets/lib/python3.6/site-packages/torch/nn/functional.py:1340: UserWarning: nn.functional.tanh is deprecated. Use torch.tanh instead.\n",
      "  warnings.warn(\"nn.functional.tanh is deprecated. Use torch.tanh instead.\")\n",
      "/home/creddy/anaconda3/envs/sets/lib/python3.6/site-packages/torch/nn/functional.py:1351: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\n",
      "  warnings.warn(\"nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\")\n"
     ]
    }
   ],
   "source": [
    "interpolation = interpolate_mini(experiment, x, verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "vutils.save_image(interpolation.cpu().data,\n",
    "                  f\"./custom_interpolation.png\",\n",
    "                  normalize=False,\n",
    "                  nrow=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fd10cc6b160>"
      ]
     },
     "execution_count": 22,
     "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": [
    "k = np.zeros([64,64,3])\n",
    "k[:,:,0] = t[0,:,:]\n",
    "k[:,:,1] = t[1,:,:]\n",
    "k[:,:,2] = t[2,:,:]\n",
    "plt.imshow(k)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}