Example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Real Time Multi Style Transfer in Pytorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training Example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training arguments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_images = \"../dataset/train2017/train2017\"\n",
    "styles_dir = \"styles\"\n",
    "num_workers = 2\n",
    "model_dir=\"new_style\"\n",
    "eval_image_dir = \"../bird.jpg\"\n",
    "eval_dir = \"temp_eval\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from trainstylenet import train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Namespace:\n",
    "    def __init__(self, **kwargs):\n",
    "        self.__dict__.update(kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "args = Namespace(content_factor=1,style_factor=1000000,lr=0.0001,eval_dir=None,epochs=2,batch_size=8, dataset_images=dataset_images, styles_dir=styles_dir,num_workers=num_workers,model_dir=model_dir,eval_image_dir=eval_image_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "train(args)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Alternatively training in command line\n",
    "```\n",
    "python multistyle_transfer_net.py --dataset-images {dataset_images} --styles-dir {styles_dir} --num-workers {num_workers} --model-dir {model_dir} --eval-image-dir {eval_image_dir} \n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Inference Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torchvision import transforms\n",
    "from inference.Inferencer import Inferencer\n",
    "from models.PasticheModel import PasticheModel\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "num_styles = 16\n",
    "image_size =512\n",
    "model_save_dir = \"style16/pastichemodel-FINAL.pth\"\n",
    "\n",
    "pastichemodel = PasticheModel(num_styles)\n",
    "\n",
    "inference = Inferencer(pastichemodel,device,image_size)\n",
    "inference.load_model_weights(model_save_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_image_path = \"../profile.jpg\"\n",
    "im = Image.open(example_image_path).convert('RGB')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%time inference.eval_image(im, 0,1,1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}