gaugan_to_onnx.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from options.test_options import TestOptions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = {'name': 'cityscapes_pretrained', 'gpu_ids': [0], 'checkpoints_dir': './checkpoints', 'model': 'pix2pix', \n",
    "     'norm_G': 'spectralspadesyncbatch3x3', 'norm_D': 'spectralinstance', 'norm_E': 'spectralinstance', \n",
    "     'phase': 'test', 'batchSize': 16, 'preprocess_mode': 'fixed', 'load_size': 512, 'crop_size': 512, 'aspect_ratio': 2.0, \n",
    "     'label_nc': 35, 'contain_dontcare_label': False, 'output_nc': 3, 'dataroot': './cityscapes', \n",
    "     'dataset_mode': 'cityscapes', 'serial_batches': True, 'no_flip': True, 'nThreads': 0, \n",
    "     'max_dataset_size': 9223372036854775807, 'load_from_opt_file': False, 'cache_filelist_write': False, \n",
    "     'cache_filelist_read': False, 'display_winsize': 512, 'netG': 'spade', 'ngf': 64, 'init_type': 'xavier', \n",
    "     'init_variance': 0.02, 'z_dim': 256, 'no_instance': False, 'nef': 16, 'use_vae': False, 'results_dir': './results/', \n",
    "     'which_epoch': 'latest', 'how_many': float('inf'), 'num_upsampling_layers': 'more', 'no_pairing_check': False, 'isTrain': False, \n",
    "     'semantic_nc': 36}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import argparse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Namespace(aspect_ratio=2.0, batchSize=16, cache_filelist_read=False, cache_filelist_write=False, checkpoints_dir='./checkpoints', contain_dontcare_label=False, crop_size=512, dataroot='./cityscapes', dataset_mode='cityscapes', display_winsize=512, gpu_ids=[0], how_many=inf, init_type='xavier', init_variance=0.02, isTrain=False, label_nc=35, load_from_opt_file=False, load_size=512, max_dataset_size=9223372036854775807, model='pix2pix', nThreads=0, name='cityscapes_pretrained', nef=16, netG='spade', ngf=64, no_flip=True, no_instance=False, no_pairing_check=False, norm_D='spectralinstance', norm_E='spectralinstance', norm_G='spectralspadesyncbatch3x3', num_upsampling_layers='more', output_nc=3, phase='test', preprocess_mode='fixed', results_dir='./results/', semantic_nc=36, serial_batches=True, use_vae=False, which_epoch='latest', z_dim=256)\n"
     ]
    }
   ],
   "source": [
    "opt = argparse.Namespace()\n",
    "opt.__dict__ = d\n",
    "print(opt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import importlib\n",
    "import models.networks.generator\n",
    "importlib.reload(models.networks.generator)\n",
    "from models.networks.generator import SPADEGeneratorOnnx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator = SPADEGeneratorOnnx(opt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "pretrained = torch.load('checkpoints/cityscapes_pretrained/latest_net_G.pth')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "generator.load_state_dict(pretrained)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "def remove_all_spectral_norm(item):\n",
    "    if isinstance(item, nn.Module):\n",
    "        try:\n",
    "            nn.utils.remove_spectral_norm(item)\n",
    "        except Exception:\n",
    "            pass\n",
    "        \n",
    "        for child in item.children():  \n",
    "            remove_all_spectral_norm(child)\n",
    "\n",
    "    if isinstance(item, nn.ModuleList):\n",
    "        for module in item:\n",
    "            remove_all_spectral_norm(module)\n",
    "\n",
    "    if isinstance(item, nn.Sequential):\n",
    "        modules = item.children()\n",
    "        for module in modules:\n",
    "            remove_all_spectral_norm(module)\n",
    "remove_all_spectral_norm(generator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\james\\stylegan2\\lib\\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"
     ]
    }
   ],
   "source": [
    "from PIL import Image\n",
    "from torchvision.transforms import transforms\n",
    "generator.train(False)\n",
    "generator.eval()\n",
    "with torch.no_grad():\n",
    "    device = torch.device('cuda')\n",
    "    generator.to(device)\n",
    "    image = Image.open('test_input.png')\n",
    "    image = image.resize((512, 256), Image.NEAREST)\n",
    "    transform_image = transforms.Compose([transforms.ToTensor()])\n",
    "    image = transform_image(image).unsqueeze(0) * 255.0\n",
    "    image = image.long()\n",
    "    image[image == 255] = 35\n",
    "    x = torch.FloatTensor(1, 35, 256, 512)\n",
    "    x = x.scatter_(1, image, 1.0).to(device)\n",
    "    image = Image.open('./instance_input.png')\n",
    "    image = image.resize((512, 256), Image.NEAREST)\n",
    "    t = transform_image(image).unsqueeze(0)\n",
    "    edge = torch.ByteTensor(t.size()).zero_()\n",
    "    edge[:, :, :, 1:] = edge[:, :, :, 1:] | (t[:, :, :, 1:] != t[:, :, :, :-1]).byte()\n",
    "    edge[:, :, :, :-1] = edge[:, :, :, :-1] | (t[:, :, :, 1:] != t[:, :, :, :-1]).byte()\n",
    "    edge[:, :, 1:, :] = edge[:, :, 1:, :] | (t[:, :, 1:, :] != t[:, :, :-1, :]).byte()\n",
    "    edge[:, :, :-1, :] = edge[:, :, :-1, :] | (t[:, :, 1:, :] != t[:, :, :-1, :]).byte()\n",
    "    edge = edge.to(device).float()\n",
    "    x = torch.cat((x, edge), dim=1)\n",
    "    z = torch.nn.functional.interpolate(x, size=(4, 8))\n",
    "    y = generator(x, z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 3, 256, 512])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x244ef5e4b38>"
      ]
     },
     "execution_count": 11,
     "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": [
    "import numpy as np\n",
    "from skimage import io\n",
    "image = np.transpose(y[0].detach().cpu().numpy(), (1, 2, 0))\n",
    "image -= image.min()\n",
    "image /= image.max()\n",
    "io.imshow(image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\james\\stylegan2\\lib\\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",
      "c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\onnx\\symbolic_helper.py:198: UserWarning: You are trying to export the model with onnx:Resize for ONNX opset version 10. This operator might cause results to not match the expected results by PyTorch.\n",
      "ONNX's Upsample/Resize operator did not match Pytorch's Interpolation until opset 11. Attributes to determine how to transform the input were added in onnx:Resize in opset 11 to support Pytorch's behavior (like coordinate_transformation_mode and nearest_mode).\n",
      "We recommend using opset 11 and above for models using this operator. \n",
      "  \"\" + str(_export_onnx_opset_version) + \". \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "graph(%input.3 : Float(1, 36, 256, 512),\n",
      "      %input.1 : Float(1, 36, 4, 8),\n",
      "      %fc.weight : Float(1024, 36, 3, 3),\n",
      "      %fc.bias : Float(1024),\n",
      "      %head_0.conv_0.bias : Float(1024),\n",
      "      %head_0.conv_0.weight : Float(1024, 1024, 3, 3),\n",
      "      %head_0.conv_1.bias : Float(1024),\n",
      "      %head_0.conv_1.weight : Float(1024, 1024, 3, 3),\n",
      "      %head_0.norm_0.param_free_norm.running_mean : Float(1024),\n",
      "      %head_0.norm_0.param_free_norm.running_var : Float(1024),\n",
      "      %head_0.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %head_0.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %head_0.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %head_0.norm_0.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %head_0.norm_0.mlp_gamma.bias : Float(1024),\n",
      "      %head_0.norm_0.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %head_0.norm_0.mlp_beta.bias : Float(1024),\n",
      "      %head_0.norm_1.param_free_norm.running_mean : Float(1024),\n",
      "      %head_0.norm_1.param_free_norm.running_var : Float(1024),\n",
      "      %head_0.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %head_0.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %head_0.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %head_0.norm_1.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %head_0.norm_1.mlp_gamma.bias : Float(1024),\n",
      "      %head_0.norm_1.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %head_0.norm_1.mlp_beta.bias : Float(1024),\n",
      "      %G_middle_0.conv_0.bias : Float(1024),\n",
      "      %G_middle_0.conv_0.weight : Float(1024, 1024, 3, 3),\n",
      "      %G_middle_0.conv_1.bias : Float(1024),\n",
      "      %G_middle_0.conv_1.weight : Float(1024, 1024, 3, 3),\n",
      "      %G_middle_0.norm_0.param_free_norm.running_mean : Float(1024),\n",
      "      %G_middle_0.norm_0.param_free_norm.running_var : Float(1024),\n",
      "      %G_middle_0.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %G_middle_0.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %G_middle_0.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %G_middle_0.norm_0.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_0.norm_0.mlp_gamma.bias : Float(1024),\n",
      "      %G_middle_0.norm_0.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_0.norm_0.mlp_beta.bias : Float(1024),\n",
      "      %G_middle_0.norm_1.param_free_norm.running_mean : Float(1024),\n",
      "      %G_middle_0.norm_1.param_free_norm.running_var : Float(1024),\n",
      "      %G_middle_0.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %G_middle_0.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %G_middle_0.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %G_middle_0.norm_1.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_0.norm_1.mlp_gamma.bias : Float(1024),\n",
      "      %G_middle_0.norm_1.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_0.norm_1.mlp_beta.bias : Float(1024),\n",
      "      %G_middle_1.conv_0.bias : Float(1024),\n",
      "      %G_middle_1.conv_0.weight : Float(1024, 1024, 3, 3),\n",
      "      %G_middle_1.conv_1.bias : Float(1024),\n",
      "      %G_middle_1.conv_1.weight : Float(1024, 1024, 3, 3),\n",
      "      %G_middle_1.norm_0.param_free_norm.running_mean : Float(1024),\n",
      "      %G_middle_1.norm_0.param_free_norm.running_var : Float(1024),\n",
      "      %G_middle_1.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %G_middle_1.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %G_middle_1.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %G_middle_1.norm_0.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_1.norm_0.mlp_gamma.bias : Float(1024),\n",
      "      %G_middle_1.norm_0.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_1.norm_0.mlp_beta.bias : Float(1024),\n",
      "      %G_middle_1.norm_1.param_free_norm.running_mean : Float(1024),\n",
      "      %G_middle_1.norm_1.param_free_norm.running_var : Float(1024),\n",
      "      %G_middle_1.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %G_middle_1.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %G_middle_1.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %G_middle_1.norm_1.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_1.norm_1.mlp_gamma.bias : Float(1024),\n",
      "      %G_middle_1.norm_1.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %G_middle_1.norm_1.mlp_beta.bias : Float(1024),\n",
      "      %up_0.conv_0.bias : Float(512),\n",
      "      %up_0.conv_0.weight : Float(512, 1024, 3, 3),\n",
      "      %up_0.conv_1.bias : Float(512),\n",
      "      %up_0.conv_1.weight : Float(512, 512, 3, 3),\n",
      "      %up_0.conv_s.weight : Float(512, 1024, 1, 1),\n",
      "      %up_0.norm_0.param_free_norm.running_mean : Float(1024),\n",
      "      %up_0.norm_0.param_free_norm.running_var : Float(1024),\n",
      "      %up_0.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_0.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_0.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %up_0.norm_0.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %up_0.norm_0.mlp_gamma.bias : Float(1024),\n",
      "      %up_0.norm_0.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %up_0.norm_0.mlp_beta.bias : Float(1024),\n",
      "      %up_0.norm_1.param_free_norm.running_mean : Float(512),\n",
      "      %up_0.norm_1.param_free_norm.running_var : Float(512),\n",
      "      %up_0.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_0.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_0.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %up_0.norm_1.mlp_gamma.weight : Float(512, 128, 3, 3),\n",
      "      %up_0.norm_1.mlp_gamma.bias : Float(512),\n",
      "      %up_0.norm_1.mlp_beta.weight : Float(512, 128, 3, 3),\n",
      "      %up_0.norm_1.mlp_beta.bias : Float(512),\n",
      "      %up_0.norm_s.param_free_norm.running_mean : Float(1024),\n",
      "      %up_0.norm_s.param_free_norm.running_var : Float(1024),\n",
      "      %up_0.norm_s.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_0.norm_s.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_0.norm_s.mlp_shared.0.bias : Float(128),\n",
      "      %up_0.norm_s.mlp_gamma.weight : Float(1024, 128, 3, 3),\n",
      "      %up_0.norm_s.mlp_gamma.bias : Float(1024),\n",
      "      %up_0.norm_s.mlp_beta.weight : Float(1024, 128, 3, 3),\n",
      "      %up_0.norm_s.mlp_beta.bias : Float(1024),\n",
      "      %up_1.conv_0.bias : Float(256),\n",
      "      %up_1.conv_0.weight : Float(256, 512, 3, 3),\n",
      "      %up_1.conv_1.bias : Float(256),\n",
      "      %up_1.conv_1.weight : Float(256, 256, 3, 3),\n",
      "      %up_1.conv_s.weight : Float(256, 512, 1, 1),\n",
      "      %up_1.norm_0.param_free_norm.running_mean : Float(512),\n",
      "      %up_1.norm_0.param_free_norm.running_var : Float(512),\n",
      "      %up_1.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_1.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_1.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %up_1.norm_0.mlp_gamma.weight : Float(512, 128, 3, 3),\n",
      "      %up_1.norm_0.mlp_gamma.bias : Float(512),\n",
      "      %up_1.norm_0.mlp_beta.weight : Float(512, 128, 3, 3),\n",
      "      %up_1.norm_0.mlp_beta.bias : Float(512),\n",
      "      %up_1.norm_1.param_free_norm.running_mean : Float(256),\n",
      "      %up_1.norm_1.param_free_norm.running_var : Float(256),\n",
      "      %up_1.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_1.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_1.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %up_1.norm_1.mlp_gamma.weight : Float(256, 128, 3, 3),\n",
      "      %up_1.norm_1.mlp_gamma.bias : Float(256),\n",
      "      %up_1.norm_1.mlp_beta.weight : Float(256, 128, 3, 3),\n",
      "      %up_1.norm_1.mlp_beta.bias : Float(256),\n",
      "      %up_1.norm_s.param_free_norm.running_mean : Float(512),\n",
      "      %up_1.norm_s.param_free_norm.running_var : Float(512),\n",
      "      %up_1.norm_s.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_1.norm_s.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_1.norm_s.mlp_shared.0.bias : Float(128),\n",
      "      %up_1.norm_s.mlp_gamma.weight : Float(512, 128, 3, 3),\n",
      "      %up_1.norm_s.mlp_gamma.bias : Float(512),\n",
      "      %up_1.norm_s.mlp_beta.weight : Float(512, 128, 3, 3),\n",
      "      %up_1.norm_s.mlp_beta.bias : Float(512),\n",
      "      %up_2.conv_0.bias : Float(128),\n",
      "      %up_2.conv_0.weight : Float(128, 256, 3, 3),\n",
      "      %up_2.conv_1.bias : Float(128),\n",
      "      %up_2.conv_1.weight : Float(128, 128, 3, 3),\n",
      "      %up_2.conv_s.weight : Float(128, 256, 1, 1),\n",
      "      %up_2.norm_0.param_free_norm.running_mean : Float(256),\n",
      "      %up_2.norm_0.param_free_norm.running_var : Float(256),\n",
      "      %up_2.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_2.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_2.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %up_2.norm_0.mlp_gamma.weight : Float(256, 128, 3, 3),\n",
      "      %up_2.norm_0.mlp_gamma.bias : Float(256),\n",
      "      %up_2.norm_0.mlp_beta.weight : Float(256, 128, 3, 3),\n",
      "      %up_2.norm_0.mlp_beta.bias : Float(256),\n",
      "      %up_2.norm_1.param_free_norm.running_mean : Float(128),\n",
      "      %up_2.norm_1.param_free_norm.running_var : Float(128),\n",
      "      %up_2.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_2.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_2.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %up_2.norm_1.mlp_gamma.weight : Float(128, 128, 3, 3),\n",
      "      %up_2.norm_1.mlp_gamma.bias : Float(128),\n",
      "      %up_2.norm_1.mlp_beta.weight : Float(128, 128, 3, 3),\n",
      "      %up_2.norm_1.mlp_beta.bias : Float(128),\n",
      "      %up_2.norm_s.param_free_norm.running_mean : Float(256),\n",
      "      %up_2.norm_s.param_free_norm.running_var : Float(256),\n",
      "      %up_2.norm_s.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_2.norm_s.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_2.norm_s.mlp_shared.0.bias : Float(128),\n",
      "      %up_2.norm_s.mlp_gamma.weight : Float(256, 128, 3, 3),\n",
      "      %up_2.norm_s.mlp_gamma.bias : Float(256),\n",
      "      %up_2.norm_s.mlp_beta.weight : Float(256, 128, 3, 3),\n",
      "      %up_2.norm_s.mlp_beta.bias : Float(256),\n",
      "      %up_3.conv_0.bias : Float(64),\n",
      "      %up_3.conv_0.weight : Float(64, 128, 3, 3),\n",
      "      %up_3.conv_1.bias : Float(64),\n",
      "      %up_3.conv_1.weight : Float(64, 64, 3, 3),\n",
      "      %up_3.conv_s.weight : Float(64, 128, 1, 1),\n",
      "      %up_3.norm_0.param_free_norm.running_mean : Float(128),\n",
      "      %up_3.norm_0.param_free_norm.running_var : Float(128),\n",
      "      %up_3.norm_0.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_3.norm_0.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_3.norm_0.mlp_shared.0.bias : Float(128),\n",
      "      %up_3.norm_0.mlp_gamma.weight : Float(128, 128, 3, 3),\n",
      "      %up_3.norm_0.mlp_gamma.bias : Float(128),\n",
      "      %up_3.norm_0.mlp_beta.weight : Float(128, 128, 3, 3),\n",
      "      %up_3.norm_0.mlp_beta.bias : Float(128),\n",
      "      %up_3.norm_1.param_free_norm.running_mean : Float(64),\n",
      "      %up_3.norm_1.param_free_norm.running_var : Float(64),\n",
      "      %up_3.norm_1.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_3.norm_1.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_3.norm_1.mlp_shared.0.bias : Float(128),\n",
      "      %up_3.norm_1.mlp_gamma.weight : Float(64, 128, 3, 3),\n",
      "      %up_3.norm_1.mlp_gamma.bias : Float(64),\n",
      "      %up_3.norm_1.mlp_beta.weight : Float(64, 128, 3, 3),\n",
      "      %up_3.norm_1.mlp_beta.bias : Float(64),\n",
      "      %up_3.norm_s.param_free_norm.running_mean : Float(128),\n",
      "      %up_3.norm_s.param_free_norm.running_var : Float(128),\n",
      "      %up_3.norm_s.param_free_norm.num_batches_tracked : Long(),\n",
      "      %up_3.norm_s.mlp_shared.0.weight : Float(128, 36, 3, 3),\n",
      "      %up_3.norm_s.mlp_shared.0.bias : Float(128),\n",
      "      %up_3.norm_s.mlp_gamma.weight : Float(128, 128, 3, 3),\n",
      "      %up_3.norm_s.mlp_gamma.bias : Float(128),\n",
      "      %up_3.norm_s.mlp_beta.weight : Float(128, 128, 3, 3),\n",
      "      %up_3.norm_s.mlp_beta.bias : Float(128),\n",
      "      %conv_img.weight : Float(3, 64, 3, 3),\n",
      "      %conv_img.bias : Float(3)):\n",
      "  %200 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%input.1, %fc.weight, %fc.bias), scope: SPADEGeneratorOnnx/Conv2d[fc] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %201 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %202 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %203 : Float(1, 1024, 4, 8) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%200, %201, %202, %head_0.norm_0.param_free_norm.running_mean, %head_0.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %204 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0156  0.0156 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %205 : Float(1, 36, 4, 8) = onnx::Resize[mode=\"nearest\"](%input.3, %204), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %206 : Float(1, 128, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%205, %head_0.norm_0.mlp_shared.0.weight, %head_0.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %207 : Float(1, 128, 4, 8) = onnx::Relu(%206), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %208 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%207, %head_0.norm_0.mlp_gamma.weight, %head_0.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %209 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%207, %head_0.norm_0.mlp_beta.weight, %head_0.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %210 : Float() = onnx::Constant[value={1}]()\n",
      "  %211 : Float(1, 1024, 4, 8) = onnx::Add(%208, %210)\n",
      "  %212 : Float(1, 1024, 4, 8) = onnx::Mul(%203, %211), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %213 : Float(1, 1024, 4, 8) = onnx::Add(%212, %209), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %214 : Float(1, 1024, 4, 8) = onnx::LeakyRelu[alpha=0.2](%213), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %215 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%214, %head_0.conv_0.weight, %head_0.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %216 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %217 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %218 : Float(1, 1024, 4, 8) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%215, %216, %217, %head_0.norm_1.param_free_norm.running_mean, %head_0.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %219 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0156  0.0156 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %220 : Float(1, 36, 4, 8) = onnx::Resize[mode=\"nearest\"](%input.3, %219), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %221 : Float(1, 128, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%220, %head_0.norm_1.mlp_shared.0.weight, %head_0.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %222 : Float(1, 128, 4, 8) = onnx::Relu(%221), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %223 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%222, %head_0.norm_1.mlp_gamma.weight, %head_0.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %224 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%222, %head_0.norm_1.mlp_beta.weight, %head_0.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %225 : Float() = onnx::Constant[value={1}]()\n",
      "  %226 : Float(1, 1024, 4, 8) = onnx::Add(%223, %225)\n",
      "  %227 : Float(1, 1024, 4, 8) = onnx::Mul(%218, %226), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %228 : Float(1, 1024, 4, 8) = onnx::Add(%227, %224), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %229 : Float(1, 1024, 4, 8) = onnx::LeakyRelu[alpha=0.2](%228), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %230 : Float(1, 1024, 4, 8) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%229, %head_0.conv_1.weight, %head_0.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %231 : Float(1, 1024, 4, 8) = onnx::Add(%200, %230), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[head_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %232 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %233 : Float(1, 1024, 8, 16) = onnx::Resize[mode=\"nearest\"](%231, %232), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %234 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %235 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %236 : Float(1, 1024, 8, 16) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%233, %234, %235, %G_middle_0.norm_0.param_free_norm.running_mean, %G_middle_0.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %237 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0313  0.0313 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %238 : Float(1, 36, 8, 16) = onnx::Resize[mode=\"nearest\"](%input.3, %237), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %239 : Float(1, 128, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%238, %G_middle_0.norm_0.mlp_shared.0.weight, %G_middle_0.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %240 : Float(1, 128, 8, 16) = onnx::Relu(%239), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %241 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%240, %G_middle_0.norm_0.mlp_gamma.weight, %G_middle_0.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %242 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%240, %G_middle_0.norm_0.mlp_beta.weight, %G_middle_0.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %243 : Float() = onnx::Constant[value={1}]()\n",
      "  %244 : Float(1, 1024, 8, 16) = onnx::Add(%241, %243)\n",
      "  %245 : Float(1, 1024, 8, 16) = onnx::Mul(%236, %244), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %246 : Float(1, 1024, 8, 16) = onnx::Add(%245, %242), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %247 : Float(1, 1024, 8, 16) = onnx::LeakyRelu[alpha=0.2](%246), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %248 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%247, %G_middle_0.conv_0.weight, %G_middle_0.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %249 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %250 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %251 : Float(1, 1024, 8, 16) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%248, %249, %250, %G_middle_0.norm_1.param_free_norm.running_mean, %G_middle_0.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %252 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0313  0.0313 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %253 : Float(1, 36, 8, 16) = onnx::Resize[mode=\"nearest\"](%input.3, %252), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %254 : Float(1, 128, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%253, %G_middle_0.norm_1.mlp_shared.0.weight, %G_middle_0.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %255 : Float(1, 128, 8, 16) = onnx::Relu(%254), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %256 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%255, %G_middle_0.norm_1.mlp_gamma.weight, %G_middle_0.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %257 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%255, %G_middle_0.norm_1.mlp_beta.weight, %G_middle_0.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %258 : Float() = onnx::Constant[value={1}]()\n",
      "  %259 : Float(1, 1024, 8, 16) = onnx::Add(%256, %258)\n",
      "  %260 : Float(1, 1024, 8, 16) = onnx::Mul(%251, %259), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %261 : Float(1, 1024, 8, 16) = onnx::Add(%260, %257), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %262 : Float(1, 1024, 8, 16) = onnx::LeakyRelu[alpha=0.2](%261), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %263 : Float(1, 1024, 8, 16) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%262, %G_middle_0.conv_1.weight, %G_middle_0.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %264 : Float(1, 1024, 8, 16) = onnx::Add(%233, %263), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %265 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %266 : Float(1, 1024, 16, 32) = onnx::Resize[mode=\"nearest\"](%264, %265), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %267 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %268 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %269 : Float(1, 1024, 16, 32) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%266, %267, %268, %G_middle_1.norm_0.param_free_norm.running_mean, %G_middle_1.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %270 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0625  0.0625 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %271 : Float(1, 36, 16, 32) = onnx::Resize[mode=\"nearest\"](%input.3, %270), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %272 : Float(1, 128, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%271, %G_middle_1.norm_0.mlp_shared.0.weight, %G_middle_1.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %273 : Float(1, 128, 16, 32) = onnx::Relu(%272), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %274 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%273, %G_middle_1.norm_0.mlp_gamma.weight, %G_middle_1.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %275 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%273, %G_middle_1.norm_0.mlp_beta.weight, %G_middle_1.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %276 : Float() = onnx::Constant[value={1}]()\n",
      "  %277 : Float(1, 1024, 16, 32) = onnx::Add(%274, %276)\n",
      "  %278 : Float(1, 1024, 16, 32) = onnx::Mul(%269, %277), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %279 : Float(1, 1024, 16, 32) = onnx::Add(%278, %275), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %280 : Float(1, 1024, 16, 32) = onnx::LeakyRelu[alpha=0.2](%279), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %281 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%280, %G_middle_1.conv_0.weight, %G_middle_1.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %282 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %283 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %284 : Float(1, 1024, 16, 32) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%281, %282, %283, %G_middle_1.norm_1.param_free_norm.running_mean, %G_middle_1.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %285 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.0625  0.0625 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %286 : Float(1, 36, 16, 32) = onnx::Resize[mode=\"nearest\"](%input.3, %285), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %287 : Float(1, 128, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%286, %G_middle_1.norm_1.mlp_shared.0.weight, %G_middle_1.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %288 : Float(1, 128, 16, 32) = onnx::Relu(%287), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %289 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%288, %G_middle_1.norm_1.mlp_gamma.weight, %G_middle_1.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %290 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%288, %G_middle_1.norm_1.mlp_beta.weight, %G_middle_1.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %291 : Float() = onnx::Constant[value={1}]()\n",
      "  %292 : Float(1, 1024, 16, 32) = onnx::Add(%289, %291)\n",
      "  %293 : Float(1, 1024, 16, 32) = onnx::Mul(%284, %292), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %294 : Float(1, 1024, 16, 32) = onnx::Add(%293, %290), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %295 : Float(1, 1024, 16, 32) = onnx::LeakyRelu[alpha=0.2](%294), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %296 : Float(1, 1024, 16, 32) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%295, %G_middle_1.conv_1.weight, %G_middle_1.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %297 : Float(1, 1024, 16, 32) = onnx::Add(%266, %296), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[G_middle_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %298 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %299 : Float(1, 1024, 32, 64) = onnx::Resize[mode=\"nearest\"](%297, %298), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %300 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %301 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %302 : Float(1, 1024, 32, 64) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%299, %300, %301, %up_0.norm_s.param_free_norm.running_mean, %up_0.norm_s.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %303 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.1250  0.1250 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Upsample[interpolate]\n",
      "  %304 : Float(1, 36, 32, 64) = onnx::Resize[mode=\"nearest\"](%input.3, %303), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %305 : Float(1, 128, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%304, %up_0.norm_s.mlp_shared.0.weight, %up_0.norm_s.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %306 : Float(1, 128, 32, 64) = onnx::Relu(%305), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %307 : Float(1, 1024, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%306, %up_0.norm_s.mlp_gamma.weight, %up_0.norm_s.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %308 : Float(1, 1024, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%306, %up_0.norm_s.mlp_beta.weight, %up_0.norm_s.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %309 : Float() = onnx::Constant[value={1}]()\n",
      "  %310 : Float(1, 1024, 32, 64) = onnx::Add(%307, %309)\n",
      "  %311 : Float(1, 1024, 32, 64) = onnx::Mul(%302, %310), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %312 : Float(1, 1024, 32, 64) = onnx::Add(%311, %308), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %313 : Float(1, 512, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%312, %up_0.conv_s.weight), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/Conv2d[conv_s] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %314 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %315 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %316 : Float(1, 1024, 32, 64) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%299, %314, %315, %up_0.norm_0.param_free_norm.running_mean, %up_0.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %317 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.1250  0.1250 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %318 : Float(1, 36, 32, 64) = onnx::Resize[mode=\"nearest\"](%input.3, %317), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %319 : Float(1, 128, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%318, %up_0.norm_0.mlp_shared.0.weight, %up_0.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %320 : Float(1, 128, 32, 64) = onnx::Relu(%319), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %321 : Float(1, 1024, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%320, %up_0.norm_0.mlp_gamma.weight, %up_0.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %322 : Float(1, 1024, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%320, %up_0.norm_0.mlp_beta.weight, %up_0.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %323 : Float() = onnx::Constant[value={1}]()\n",
      "  %324 : Float(1, 1024, 32, 64) = onnx::Add(%321, %323)\n",
      "  %325 : Float(1, 1024, 32, 64) = onnx::Mul(%316, %324), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %326 : Float(1, 1024, 32, 64) = onnx::Add(%325, %322), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %327 : Float(1, 1024, 32, 64) = onnx::LeakyRelu[alpha=0.2](%326), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %328 : Float(1, 512, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%327, %up_0.conv_0.weight, %up_0.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %329 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %330 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %331 : Float(1, 512, 32, 64) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%328, %329, %330, %up_0.norm_1.param_free_norm.running_mean, %up_0.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %332 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.1250  0.1250 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %333 : Float(1, 36, 32, 64) = onnx::Resize[mode=\"nearest\"](%input.3, %332), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %334 : Float(1, 128, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%333, %up_0.norm_1.mlp_shared.0.weight, %up_0.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %335 : Float(1, 128, 32, 64) = onnx::Relu(%334), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %336 : Float(1, 512, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%335, %up_0.norm_1.mlp_gamma.weight, %up_0.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %337 : Float(1, 512, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%335, %up_0.norm_1.mlp_beta.weight, %up_0.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %338 : Float() = onnx::Constant[value={1}]()\n",
      "  %339 : Float(1, 512, 32, 64) = onnx::Add(%336, %338)\n",
      "  %340 : Float(1, 512, 32, 64) = onnx::Mul(%331, %339), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %341 : Float(1, 512, 32, 64) = onnx::Add(%340, %337), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %342 : Float(1, 512, 32, 64) = onnx::LeakyRelu[alpha=0.2](%341), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %343 : Float(1, 512, 32, 64) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%342, %up_0.conv_1.weight, %up_0.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %344 : Float(1, 512, 32, 64) = onnx::Add(%313, %343), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %345 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %346 : Float(1, 512, 64, 128) = onnx::Resize[mode=\"nearest\"](%344, %345), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %347 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %348 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %349 : Float(1, 512, 64, 128) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%346, %347, %348, %up_1.norm_s.param_free_norm.running_mean, %up_1.norm_s.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %350 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.2500  0.2500 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Upsample[interpolate]\n",
      "  %351 : Float(1, 36, 64, 128) = onnx::Resize[mode=\"nearest\"](%input.3, %350), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %352 : Float(1, 128, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%351, %up_1.norm_s.mlp_shared.0.weight, %up_1.norm_s.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %353 : Float(1, 128, 64, 128) = onnx::Relu(%352), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %354 : Float(1, 512, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%353, %up_1.norm_s.mlp_gamma.weight, %up_1.norm_s.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %355 : Float(1, 512, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%353, %up_1.norm_s.mlp_beta.weight, %up_1.norm_s.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %356 : Float() = onnx::Constant[value={1}]()\n",
      "  %357 : Float(1, 512, 64, 128) = onnx::Add(%354, %356)\n",
      "  %358 : Float(1, 512, 64, 128) = onnx::Mul(%349, %357), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %359 : Float(1, 512, 64, 128) = onnx::Add(%358, %355), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %360 : Float(1, 256, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%359, %up_1.conv_s.weight), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/Conv2d[conv_s] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %361 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %362 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %363 : Float(1, 512, 64, 128) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%346, %361, %362, %up_1.norm_0.param_free_norm.running_mean, %up_1.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %364 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.2500  0.2500 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %365 : Float(1, 36, 64, 128) = onnx::Resize[mode=\"nearest\"](%input.3, %364), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %366 : Float(1, 128, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%365, %up_1.norm_0.mlp_shared.0.weight, %up_1.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %367 : Float(1, 128, 64, 128) = onnx::Relu(%366), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %368 : Float(1, 512, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%367, %up_1.norm_0.mlp_gamma.weight, %up_1.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %369 : Float(1, 512, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%367, %up_1.norm_0.mlp_beta.weight, %up_1.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %370 : Float() = onnx::Constant[value={1}]()\n",
      "  %371 : Float(1, 512, 64, 128) = onnx::Add(%368, %370)\n",
      "  %372 : Float(1, 512, 64, 128) = onnx::Mul(%363, %371), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %373 : Float(1, 512, 64, 128) = onnx::Add(%372, %369), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %374 : Float(1, 512, 64, 128) = onnx::LeakyRelu[alpha=0.2](%373), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %375 : Float(1, 256, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%374, %up_1.conv_0.weight, %up_1.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %376 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %377 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %378 : Float(1, 256, 64, 128) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%375, %376, %377, %up_1.norm_1.param_free_norm.running_mean, %up_1.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %379 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.2500  0.2500 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %380 : Float(1, 36, 64, 128) = onnx::Resize[mode=\"nearest\"](%input.3, %379), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %381 : Float(1, 128, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%380, %up_1.norm_1.mlp_shared.0.weight, %up_1.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %382 : Float(1, 128, 64, 128) = onnx::Relu(%381), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %383 : Float(1, 256, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%382, %up_1.norm_1.mlp_gamma.weight, %up_1.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %384 : Float(1, 256, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%382, %up_1.norm_1.mlp_beta.weight, %up_1.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %385 : Float() = onnx::Constant[value={1}]()\n",
      "  %386 : Float(1, 256, 64, 128) = onnx::Add(%383, %385)\n",
      "  %387 : Float(1, 256, 64, 128) = onnx::Mul(%378, %386), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %388 : Float(1, 256, 64, 128) = onnx::Add(%387, %384), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %389 : Float(1, 256, 64, 128) = onnx::LeakyRelu[alpha=0.2](%388), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %390 : Float(1, 256, 64, 128) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%389, %up_1.conv_1.weight, %up_1.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %391 : Float(1, 256, 64, 128) = onnx::Add(%360, %390), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %392 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %393 : Float(1, 256, 128, 256) = onnx::Resize[mode=\"nearest\"](%391, %392), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %394 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %395 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %396 : Float(1, 256, 128, 256) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%393, %394, %395, %up_2.norm_s.param_free_norm.running_mean, %up_2.norm_s.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %397 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.5000  0.5000 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Upsample[interpolate]\n",
      "  %398 : Float(1, 36, 128, 256) = onnx::Resize[mode=\"nearest\"](%input.3, %397), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %399 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%398, %up_2.norm_s.mlp_shared.0.weight, %up_2.norm_s.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %400 : Float(1, 128, 128, 256) = onnx::Relu(%399), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %401 : Float(1, 256, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%400, %up_2.norm_s.mlp_gamma.weight, %up_2.norm_s.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %402 : Float(1, 256, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%400, %up_2.norm_s.mlp_beta.weight, %up_2.norm_s.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %403 : Float() = onnx::Constant[value={1}]()\n",
      "  %404 : Float(1, 256, 128, 256) = onnx::Add(%401, %403)\n",
      "  %405 : Float(1, 256, 128, 256) = onnx::Mul(%396, %404), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %406 : Float(1, 256, 128, 256) = onnx::Add(%405, %402), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %407 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%406, %up_2.conv_s.weight), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/Conv2d[conv_s] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %408 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %409 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %410 : Float(1, 256, 128, 256) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%393, %408, %409, %up_2.norm_0.param_free_norm.running_mean, %up_2.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %411 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.5000  0.5000 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %412 : Float(1, 36, 128, 256) = onnx::Resize[mode=\"nearest\"](%input.3, %411), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %413 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%412, %up_2.norm_0.mlp_shared.0.weight, %up_2.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %414 : Float(1, 128, 128, 256) = onnx::Relu(%413), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %415 : Float(1, 256, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%414, %up_2.norm_0.mlp_gamma.weight, %up_2.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %416 : Float(1, 256, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%414, %up_2.norm_0.mlp_beta.weight, %up_2.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %417 : Float() = onnx::Constant[value={1}]()\n",
      "  %418 : Float(1, 256, 128, 256) = onnx::Add(%415, %417)\n",
      "  %419 : Float(1, 256, 128, 256) = onnx::Mul(%410, %418), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %420 : Float(1, 256, 128, 256) = onnx::Add(%419, %416), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %421 : Float(1, 256, 128, 256) = onnx::LeakyRelu[alpha=0.2](%420), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %422 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%421, %up_2.conv_0.weight, %up_2.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %423 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %424 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %425 : Float(1, 128, 128, 256) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%422, %423, %424, %up_2.norm_1.param_free_norm.running_mean, %up_2.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %426 : Tensor = onnx::Constant[value= 1.0000  1.0000  0.5000  0.5000 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %427 : Float(1, 36, 128, 256) = onnx::Resize[mode=\"nearest\"](%input.3, %426), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %428 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%427, %up_2.norm_1.mlp_shared.0.weight, %up_2.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %429 : Float(1, 128, 128, 256) = onnx::Relu(%428), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %430 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%429, %up_2.norm_1.mlp_gamma.weight, %up_2.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %431 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%429, %up_2.norm_1.mlp_beta.weight, %up_2.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %432 : Float() = onnx::Constant[value={1}]()\n",
      "  %433 : Float(1, 128, 128, 256) = onnx::Add(%430, %432)\n",
      "  %434 : Float(1, 128, 128, 256) = onnx::Mul(%425, %433), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %435 : Float(1, 128, 128, 256) = onnx::Add(%434, %431), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %436 : Float(1, 128, 128, 256) = onnx::LeakyRelu[alpha=0.2](%435), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %437 : Float(1, 128, 128, 256) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%436, %up_2.conv_1.weight, %up_2.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %438 : Float(1, 128, 128, 256) = onnx::Add(%407, %437), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_2] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %439 : Tensor = onnx::Constant[value= 1  1  2  2 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/Upsample\n",
      "  %440 : Float(1, 128, 256, 512) = onnx::Resize[mode=\"nearest\"](%438, %439), scope: SPADEGeneratorOnnx/Upsample # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %441 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %442 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %443 : Float(1, 128, 256, 512) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%440, %441, %442, %up_3.norm_s.param_free_norm.running_mean, %up_3.norm_s.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %444 : Tensor = onnx::Constant[value= 1  1  1  1 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Upsample[interpolate]\n",
      "  %445 : Float(1, 36, 256, 512) = onnx::Resize[mode=\"nearest\"](%input.3, %444), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %446 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%445, %up_3.norm_s.mlp_shared.0.weight, %up_3.norm_s.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %447 : Float(1, 128, 256, 512) = onnx::Relu(%446), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %448 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%447, %up_3.norm_s.mlp_gamma.weight, %up_3.norm_s.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %449 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%447, %up_3.norm_s.mlp_beta.weight, %up_3.norm_s.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %450 : Float() = onnx::Constant[value={1}]()\n",
      "  %451 : Float(1, 128, 256, 512) = onnx::Add(%448, %450)\n",
      "  %452 : Float(1, 128, 256, 512) = onnx::Mul(%443, %451), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %453 : Float(1, 128, 256, 512) = onnx::Add(%452, %449), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_s] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %454 : Float(1, 64, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%453, %up_3.conv_s.weight), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/Conv2d[conv_s] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %455 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %456 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %457 : Float(1, 128, 256, 512) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%440, %455, %456, %up_3.norm_0.param_free_norm.running_mean, %up_3.norm_0.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %458 : Tensor = onnx::Constant[value= 1  1  1  1 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Upsample[interpolate]\n",
      "  %459 : Float(1, 36, 256, 512) = onnx::Resize[mode=\"nearest\"](%input.3, %458), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %460 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%459, %up_3.norm_0.mlp_shared.0.weight, %up_3.norm_0.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %461 : Float(1, 128, 256, 512) = onnx::Relu(%460), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %462 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%461, %up_3.norm_0.mlp_gamma.weight, %up_3.norm_0.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %463 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%461, %up_3.norm_0.mlp_beta.weight, %up_3.norm_0.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %464 : Float() = onnx::Constant[value={1}]()\n",
      "  %465 : Float(1, 128, 256, 512) = onnx::Add(%462, %464)\n",
      "  %466 : Float(1, 128, 256, 512) = onnx::Mul(%457, %465), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %467 : Float(1, 128, 256, 512) = onnx::Add(%466, %463), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_0] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %468 : Float(1, 128, 256, 512) = onnx::LeakyRelu[alpha=0.2](%467), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %469 : Float(1, 64, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%468, %up_3.conv_0.weight, %up_3.conv_0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/Conv2d[conv_0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %470 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %471 : Tensor = onnx::Constant[value=<Tensor>](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm]\n",
      "  %472 : Float(1, 64, 256, 512) = onnx::BatchNormalization[epsilon=1e-05, momentum=0.9](%469, %470, %471, %up_3.norm_1.param_free_norm.running_mean, %up_3.norm_1.param_free_norm.running_var), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/SynchronizedBatchNorm2d[param_free_norm] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1670:0\n",
      "  %473 : Tensor = onnx::Constant[value= 1  1  1  1 [ Variable[CPUFloatType]{4} ]](), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Upsample[interpolate]\n",
      "  %474 : Float(1, 36, 256, 512) = onnx::Resize[mode=\"nearest\"](%input.3, %473), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Upsample[interpolate] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:2500:0\n",
      "  %475 : Float(1, 128, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%474, %up_3.norm_1.mlp_shared.0.weight, %up_3.norm_1.mlp_shared.0.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Sequential[mlp_shared]/Conv2d[0] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %476 : Float(1, 128, 256, 512) = onnx::Relu(%475), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Sequential[mlp_shared]/ReLU[1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:914:0\n",
      "  %477 : Float(1, 64, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%476, %up_3.norm_1.mlp_gamma.weight, %up_3.norm_1.mlp_gamma.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Conv2d[mlp_gamma] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %478 : Float(1, 64, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%476, %up_3.norm_1.mlp_beta.weight, %up_3.norm_1.mlp_beta.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1]/Conv2d[mlp_beta] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %479 : Float() = onnx::Constant[value={1}]()\n",
      "  %480 : Float(1, 64, 256, 512) = onnx::Add(%477, %479)\n",
      "  %481 : Float(1, 64, 256, 512) = onnx::Mul(%472, %480), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %482 : Float(1, 64, 256, 512) = onnx::Add(%481, %478), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/SPADE[norm_1] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\normalization.py:123:0\n",
      "  %483 : Float(1, 64, 256, 512) = onnx::LeakyRelu[alpha=0.2](%482), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %484 : Float(1, 64, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%483, %up_3.conv_1.weight, %up_3.conv_1.bias), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3]/Conv2d[conv_1] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %485 : Float(1, 64, 256, 512) = onnx::Add(%454, %484), scope: SPADEGeneratorOnnx/SPADEResnetBlockOnnx[up_3] # C:\\Users\\James\\projects\\SPADE\\models\\networks\\architecture.py:111:0\n",
      "  %486 : Float(1, 64, 256, 512) = onnx::LeakyRelu[alpha=0.2](%485), scope: SPADEGeneratorOnnx # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1063:0\n",
      "  %487 : Float(1, 3, 256, 512) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%486, %conv_img.weight, %conv_img.bias), scope: SPADEGeneratorOnnx/Conv2d[conv_img] # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\modules\\conv.py:342:0\n",
      "  %488 : Float(1, 3, 256, 512) = onnx::Tanh(%487), scope: SPADEGeneratorOnnx # c:\\users\\james\\stylegan2\\lib\\site-packages\\torch\\nn\\functional.py:1341:0\n",
      "  return (%488)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "generator.train(False)\n",
    "generator.eval()\n",
    "generator.to('cpu')\n",
    "device = torch.device('cuda')\n",
    "x = x.cpu()\n",
    "z = z.cpu()\n",
    "x = (x, z)\n",
    "# import torch.onnx.symbolic_opset10 as onnx_symbolic\n",
    "# def upsample_nearest2d(g, input, output_size, *args):\n",
    "#     # Currently, TRT 5.1/6.0/7.0 ONNX Parser does not support all ONNX ops\n",
    "#     # needed to support dynamic upsampling ONNX forumlation\n",
    "#     # Here we hardcode scale=2 as a temporary workaround\n",
    "#     scales = g.op(\"Constant\", value_t=torch.tensor([1., 1., 2., 2.]))\n",
    "#     return g.op(\"Resize\", input, scales, mode_s=\"nearest\")\n",
    "\n",
    "# onnx_symbolic.upsample_nearest2d = upsample_nearest2d\n",
    "torch.onnx.export(generator,\n",
    "                  x,\n",
    "                  \"./gaugan_c_upsample.onnx\",\n",
    "                  export_params=True,\n",
    "                  opset_version=10,\n",
    "                  verbose=True\n",
    "                 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Upsample(scale_factor=0.015625, mode=nearest)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "generator.head_0.norm_0.interpolate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import onnx\n",
    "onnx_model = onnx.load('./gaugan_c_upsample.onnx')\n",
    "onnx.checker.check_model(onnx_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import onnxruntime\n",
    "import numpy as np\n",
    "options = onnxruntime.SessionOptions()\n",
    "options.log_severity_level = 0\n",
    "options.log_verbosity_level = 4\n",
    "onnx_session = onnxruntime.InferenceSession('./gaugan_c_upsample.onnx', options)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "in_name_1 = onnx_session.get_inputs()[0].name\n",
    "in_name_2 = onnx_session.get_inputs()[1].name\n",
    "from PIL import Image\n",
    "from torchvision.transforms import transforms\n",
    "\n",
    "image = Image.open('test_input.png')\n",
    "image = image.resize((512, 256), Image.NEAREST)\n",
    "transform_image = transforms.Compose([transforms.ToTensor()])\n",
    "image = transform_image(image).unsqueeze(0) * 255.0\n",
    "image = image.long()\n",
    "image[image == 255] = 35\n",
    "x = torch.FloatTensor(1, 35, 256, 512)\n",
    "x = x.scatter_(1, image, 1.0).cpu()\n",
    "image = Image.open('./instance_input.png')\n",
    "image = image.resize((512, 256), Image.NEAREST)\n",
    "t = transform_image(image).unsqueeze(0)\n",
    "edge = torch.ByteTensor(t.size()).zero_()\n",
    "edge[:, :, :, 1:] = edge[:, :, :, 1:] | (t[:, :, :, 1:] != t[:, :, :, :-1]).byte()\n",
    "edge[:, :, :, :-1] = edge[:, :, :, :-1] | (t[:, :, :, 1:] != t[:, :, :, :-1]).byte()\n",
    "edge[:, :, 1:, :] = edge[:, :, 1:, :] | (t[:, :, 1:, :] != t[:, :, :-1, :]).byte()\n",
    "edge[:, :, :-1, :] = edge[:, :, :-1, :] | (t[:, :, 1:, :] != t[:, :, :-1, :]).byte()\n",
    "edge = edge.float()\n",
    "x = torch.cat((x, edge), dim=1)\n",
    "z = torch.nn.functional.interpolate(x, size=(4, 8)).cpu().numpy()\n",
    "x = x.cpu().numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "onnx_inputs = {in_name_1: x, in_name_2: z}\n",
    "options = onnxruntime.RunOptions()\n",
    "options.log_severity_level = 0\n",
    "options.log_verbosity_level = 4\n",
    "onnx_output = onnx_session.run(None, onnx_inputs, options)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x245089636a0>"
      ]
     },
     "execution_count": 16,
     "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": [
    "from skimage import io\n",
    "onnx_output[0] -= onnx_output[0].min()\n",
    "onnx_output[0] /= onnx_output[0].max()\n",
    "io.imshow(np.transpose(onnx_output[0][0], (1, 2, 0)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 1, 1, 4])\n",
      "torch.Size([1, 1, 2, 4])\n"
     ]
    }
   ],
   "source": [
    "_a = torch.Tensor([[[[1, 2, 3, 4]]]])\n",
    "print(_a.shape)\n",
    "_a = torch.nn.functional.interpolate(_a, size=(2, 4))\n",
    "print(_a.shape)"
   ]
  },
  {
   "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": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}