setup.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1><span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"></ul></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import torchvision\n",
    "\n",
    "from model import BarlowTwins, Transform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ToTensor()"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Compose(\n",
       "    Grayscale(num_output_channels=3)\n",
       "    RandomResizedCrop(size=(224, 224), scale=(0.08, 1.0), ratio=(0.75, 1.3333), interpolation=bicubic)\n",
       "    RandomHorizontalFlip(p=0.5)\n",
       "    RandomApply(\n",
       "    p=0.8\n",
       "    ColorJitter(brightness=[0.6, 1.4], contrast=[0.6, 1.4], saturation=[0.8, 1.2], hue=[-0.1, 0.1])\n",
       ")\n",
       "    RandomGrayscale(p=0.2)\n",
       "    <model.GaussianBlur object at 0x1383de8e0>\n",
       "    <model.Solarization object at 0x1383ca8e0>\n",
       "    ToTensor()\n",
       "    Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
       ")"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Transform().transform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e379b3cb9a134b099b499a621ff82e24",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/170498071 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data/cifar-10-python.tar.gz to ./data/\n"
     ]
    }
   ],
   "source": [
    "dataset = torchvision.datasets.CIFAR10(\n",
    "  './data/', \n",
    "  train=True, \n",
    "  download=True,\n",
    "  transform=transforms.Compose([transforms.ToTensor()])\n",
    "#   transform=Transform()\n",
    "  )\n",
    "\n",
    "loader = torch.utils.data.DataLoader(\n",
    "    dataset,\n",
    "    batch_size=2,\n",
    "    num_workers=1,\n",
    "    pin_memory=True,\n",
    "#     sampler=sampler,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dataclasses import dataclass\n",
    "\n",
    "@dataclass\n",
    "class Config:\n",
    "    model='resnet50'\n",
    "    proj_dim=100\n",
    "    proj_layers=2\n",
    "    batch_size=2\n",
    "    lambda_=0.1\n",
    "    pretrained=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = BarlowTwins(Config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4b98ef6edc6d4d66b92e596c48b37a8b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m/var/folders/8w/r6kg1v9x7bbfzf9dw9gjslc80000gn/T/ipykernel_7424/3467416288.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0;31m#         optimizer.zero_grad()\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     13\u001b[0m         \u001b[0;32mwith\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcuda\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mamp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautocast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m             \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     15\u001b[0m         \u001b[0mscaler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscale\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;31m#         scaler.step(optimizer)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/dev/repos/barlow_twins_pytorch/model.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, y1, y2)\u001b[0m\n\u001b[1;32m     59\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     60\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 61\u001b[0;31m         \u001b[0mz1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprojector\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mencoder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     62\u001b[0m         \u001b[0mz2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprojector\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mencoder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     63\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1100\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1101\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1102\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1103\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/dev/repos/barlow_twins_pytorch/model.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m     27\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     28\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 29\u001b[0;31m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     30\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1100\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1101\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1102\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1103\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/models/resnet.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    683\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    684\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 685\u001b[0;31m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward_features\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    686\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mglobal_pool\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    687\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdrop_rate\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/models/resnet.py\u001b[0m in \u001b[0;36mforward_features\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    678\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayer1\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    679\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayer2\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 680\u001b[0;31m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayer3\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    681\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayer4\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    682\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1100\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1101\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1102\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1103\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/container.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    139\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    140\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 141\u001b[0;31m             \u001b[0minput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    142\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    143\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1100\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1101\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1102\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1103\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/models/resnet.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    407\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mact1\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    408\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 409\u001b[0;31m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconv2\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    410\u001b[0m         \u001b[0mx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbn2\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    411\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdrop_block\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1100\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[1;32m   1101\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[0;32m-> 1102\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1103\u001b[0m         \u001b[0;31m# Do not call functions when jit is used\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1104\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/conv.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    444\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    445\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m->\u001b[0m \u001b[0mTensor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 446\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_conv_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbias\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    447\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    448\u001b[0m \u001b[0;32mclass\u001b[0m \u001b[0mConv3d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_ConvNd\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torch/nn/modules/conv.py\u001b[0m in \u001b[0;36m_conv_forward\u001b[0;34m(self, input, weight, bias)\u001b[0m\n\u001b[1;32m    440\u001b[0m                             \u001b[0mweight\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbias\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstride\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    441\u001b[0m                             _pair(0), self.dilation, self.groups)\n\u001b[0;32m--> 442\u001b[0;31m         return F.conv2d(input, weight, bias, self.stride,\n\u001b[0m\u001b[1;32m    443\u001b[0m                         self.padding, self.dilation, self.groups)\n\u001b[1;32m    444\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "import time\n",
    "from tqdm.auto import tqdm \n",
    "\n",
    "start_time = time.time()\n",
    "scaler = torch.cuda.amp.GradScaler()\n",
    "for epoch in range(0, 1):\n",
    "#     sampler.set_epoch(epoch)\n",
    "    for step, ((y1, y2), _) in tqdm(enumerate(loader, start=epoch * len(loader))):\n",
    "#         y1 = y1.cuda(gpu, non_blocking=True)\n",
    "#         y2 = y2.cuda(gpu, non_blocking=True)\n",
    "#         adjust_learning_rate(args, optimizer, loader, step)\n",
    "#         optimizer.zero_grad()\n",
    "        with torch.cuda.amp.autocast():\n",
    "            loss = model.forward(y1, y2)\n",
    "        scaler.scale(loss).backward()\n",
    "#         scaler.step(optimizer)\n",
    "#         scaler.update()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0., 1., 2.],\n",
       "        [3., 0., 5.],\n",
       "        [6., 7., 0.]])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mask = torch.ones(3, 3, dtype=torch.float32) - torch.eye(3, dtype=torch.float32)\n",
    "torch.arange(0, 9, 1).view(3,3).mul(mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import torchvision\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
      "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./data/MNIST/raw/train-images-idx3-ubyte.gz\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "72dd24da0f234623b404ec1b2ca8c7c8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/9912422 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting ./data/MNIST/raw/train-images-idx3-ubyte.gz to ./data/MNIST/raw\n",
      "\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m/var/folders/8w/r6kg1v9x7bbfzf9dw9gjslc80000gn/T/ipykernel_7424/2353798407.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      7\u001b[0m   ])\n\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 9\u001b[0;31m train_data = torchvision.datasets.MNIST(\n\u001b[0m\u001b[1;32m     10\u001b[0m   \u001b[0;34m'./data/'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m   \u001b[0mtrain\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torchvision/datasets/mnist.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, root, train, transform, target_transform, download)\u001b[0m\n\u001b[1;32m     85\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     86\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mdownload\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 87\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdownload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     88\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     89\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_check_exists\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torchvision/datasets/mnist.py\u001b[0m in \u001b[0;36mdownload\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    174\u001b[0m                 \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    175\u001b[0m                     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Downloading {}\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0murl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 176\u001b[0;31m                     download_and_extract_archive(\n\u001b[0m\u001b[1;32m    177\u001b[0m                         \u001b[0murl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdownload_root\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mraw_folder\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    178\u001b[0m                         \u001b[0mfilename\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mfilename\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torchvision/datasets/utils.py\u001b[0m in \u001b[0;36mdownload_and_extract_archive\u001b[0;34m(url, download_root, extract_root, filename, md5, remove_finished)\u001b[0m\n\u001b[1;32m    429\u001b[0m     \u001b[0marchive\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdownload_root\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfilename\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    430\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Extracting {} to {}\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marchive\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mextract_root\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 431\u001b[0;31m     \u001b[0mextract_archive\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marchive\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mextract_root\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mremove_finished\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    432\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    433\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torchvision/datasets/utils.py\u001b[0m in \u001b[0;36mextract_archive\u001b[0;34m(from_path, to_path, remove_finished)\u001b[0m\n\u001b[1;32m    397\u001b[0m     \u001b[0msuffix\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0marchive_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcompression\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_detect_file_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfrom_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    398\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0marchive_type\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 399\u001b[0;31m         return _decompress(\n\u001b[0m\u001b[1;32m    400\u001b[0m             \u001b[0mfrom_path\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    401\u001b[0m             \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mto_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbasename\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfrom_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreplace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msuffix\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/torchvision/datasets/utils.py\u001b[0m in \u001b[0;36m_decompress\u001b[0;34m(from_path, to_path, remove_finished)\u001b[0m\n\u001b[1;32m    369\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    370\u001b[0m     \u001b[0;32mwith\u001b[0m \u001b[0mcompressed_file_opener\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfrom_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"rb\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mrfh\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mto_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"wb\"\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mwfh\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 371\u001b[0;31m         \u001b[0mwfh\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrfh\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    372\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    373\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mremove_finished\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/lib/python3.8/gzip.py\u001b[0m in \u001b[0;36mread\u001b[0;34m(self, size)\u001b[0m\n\u001b[1;32m    290\u001b[0m             \u001b[0;32mimport\u001b[0m \u001b[0merrno\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    291\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mOSError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merrno\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mEBADF\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"read() on write-only GzipFile object\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 292\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_buffer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    293\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    294\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mread1\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/lib/python3.8/gzip.py\u001b[0m in \u001b[0;36mread\u001b[0;34m(self, size)\u001b[0m\n\u001b[1;32m    485\u001b[0m             \u001b[0mbuf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mio\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDEFAULT_BUFFER_SIZE\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    486\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 487\u001b[0;31m             \u001b[0muncompress\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_decompressor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdecompress\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbuf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    488\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_decompressor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munconsumed_tail\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34mb\"\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    489\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprepend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_decompressor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munconsumed_tail\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "torch.backends.cudnn.enabled = False\n",
    "torch.manual_seed(123)\n",
    "\n",
    "mnist_transform = torchvision.transforms.Compose([\n",
    "  torchvision.transforms.ToTensor(),\n",
    "  torchvision.transforms.Normalize((0.1307,), (0.3081,))\n",
    "  ])\n",
    "\n",
    "train_data = torchvision.datasets.MNIST(\n",
    "  './data/', \n",
    "  train=True, \n",
    "  download=True,\n",
    "  transform=mnist_transform\n",
    "  )\n",
    "\n",
    "test_data = torchvision.datasets.MNIST(\n",
    "  './data/', \n",
    "  train=False, \n",
    "  download=True,\n",
    "  transform=mnist_transform\n",
    "  )\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(\n",
    "  train_data,\n",
    "  batch_size=4, \n",
    "  shuffle=True\n",
    "  )\n",
    "\n",
    "test_loader = torch.utils.data.DataLoader(\n",
    "  test_data,\n",
    "  batch_size=2, \n",
    "  shuffle=False\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13374bfd0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "images, labels = train_data[0]\n",
    "plt.imshow(images.reshape(28, 28), cmap='gray')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "examples = enumerate(train_loader)\n",
    "batch_idx, (example_data, example_targets) = next(examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torchvision import transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomResizedCropAndInterpolation(size=(224, 224), scale=(0.08, 1.0), ratio=(0.75, 1.3333), interpolation=bilinear)"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "from timm.data.auto_augment import rand_augment_transform\n",
    "tfm = rand_augment_transform(\n",
    "    config_str='rand-m9-mstd0.5',\n",
    "    hparams={'img_mean': (124, 116, 104)}\n",
    ")\n",
    "\n",
    "class Augmenter:\n",
    "    def __init__(self, tfm):\n",
    "        self.tfm_a = tfm\n",
    "        self.tfm_b = tfm\n",
    "    \n",
    "    def __call__(self, x):\n",
    "    \n",
    "        x1 = self.tfm_a(x)\n",
    "        x2 = self.tfm_b(x)\n",
    "        return x1, x2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "aug = Augmenter(tfm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch = next(iter(loader))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [],
   "source": [
    "toImg = transforms.ToPILImage()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "img_tensor = batch[0][0][1,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Image.BILINEAR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to cifar10/cifar-10-python.tar.gz\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "72a997e5b2fb448291989602555a1ed7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/170498071 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting cifar10/cifar-10-python.tar.gz to cifar10\n"
     ]
    }
   ],
   "source": [
    "from timm.data import create_dataset\n",
    "ds = create_dataset('torch/cifar10', 'cifar10', download=True, split='train[:5%]')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "img, label = ds[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13acd20a0>"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD5CAYAAADhukOtAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAe8klEQVR4nO2da4yc53Xf/2feuex9Z5dLLpdXURJlRVZiSqFVO1EV2akDRUkgGwhcu4ChAEYUBBEQA+kHwQVqF+gHp6ht+EPhgq5VK4ZrWbUtSEiE1LYcRDDsSKJu1IW6ULxIJJdcksu97+zcTj/MyKXU5//sksudpf38fwDB2efs875nnnnPvLPPf8455u4QQvz6k1tvB4QQnUHBLkQiKNiFSAQFuxCJoGAXIhEU7EIkQn41k83sDgBfA5AB+B/u/qXY7/d3533DQDF8rPh5Ltq3mKTo4Lbouci06PH40eJGj70Px/wP2yx2MjIHAGLK7KXJttyP2NHcL/4aaB2TrQenGX3Sl+ZH7NkxSzPiBvNxer6OxaVG0MlLDnYzywD8NwAfA3AcwNNm9qi7v8LmbBgo4gv/7vrw8bxJz1UshN20HA+IanWJ2uqNGj9XMfxmBACNZthHj7wqlmtQWy6jJnitlx8T/JiFYiU4nkVeastx/xvNOrXV6vw1azZJUBj3ox6+RgEAS+x4WC5wwz7G3tSrVX59NBqRdYxcw7nIa1Yl19U8X3osVMPH+/ZPTkR8uHRuAXDI3Q+7exXAgwDuWsXxhBBryGqCfSuAty/4+Xh7TAhxBbLmG3Rmdo+Z7Tez/XOLkc8lQog1ZTXBfgLA9gt+3tYeexfuvs/d97r73r7uVe0HCiFWwWqC/WkAu81sl5kVAXwKwKOXxy0hxOXmkm+17l43s3sB/B+0pLf73f3l6BwYquT9xX2RTyS7lSXwHesc+FZ3Ph/ZIb8ExcsKfNJStUpt9WbEx4j0lkV28fNkmjX5DjPqXLmI7SI3I/5XrSs43shKfE7seA2+HtbkPhpRE7oir1neuC2XjygXtcgaG/8T1skae0RnyLKwjzFlYlWfq939MQCPreYYQojOoG/QCZEICnYhEkHBLkQiKNiFSAQFuxCJ0OFvuTicJVY4l3+8EZ5jDS7VNGtc8sq6IzIOeDIDk7yaEemnWChQW925rVmLPLfI+er1sM0imVy5iMxnGU8M8iwsrwHAYiMssZ06x+Wp+Sr3cW6Oz8ucr0d/V3gdi8Zf54GebmrrLnEJrZnj11wuKqOFfeRXB1BjyVcR7U13diESQcEuRCIo2IVIBAW7EImgYBciETq6G2/uyDfIrnsW2S0mSRylLJIfn49tS0YSHUiCAQCaCFOPFQvLcT8KRb7ru/mq66htZuostZ09txA+V57vqucQSU6p80tk0bn/B4+FffTSMJ1Ty3hiU7WP7/zPTU9S24mJqeB4X4k/r8ap8BwA2DHK13FDP1/HrnysnFX4Oi5GLuEGUSBi5bZ0ZxciERTsQiSCgl2IRFCwC5EICnYhEkHBLkQirEO517A0YPkyn0HkhHqsA0eOy3LVOk9YKEZqpDUapFZYJDEFESmkGKmD9q/+zceo7Zmf/4LaTk6dC47PRyS0eoNLXseOn6G2Iyd495FSeSw4vm10F53jpX5qq+b561Lo20ht9cpccPzcxEk6p6fM5cHjc6eprUJqJQLAaD9Pa+kphBNhGrWwjAoArIlPpJOX7uxCpIKCXYhEULALkQgKdiESQcEuRCIo2IVIhFVJb2Z2FMAsgAaAurvvjf1+03JYyoXllemFHjqvQdoTDfVxeW0g43JYPlKPrRmR5ZisQevqIZ5Ft7Bwntp++vePUNvpKV6v7/Rc+HzHTvBzHRt/m9qyrj5qa2QD1NY7MBIcL/Tw4+W7eBZdKdKSqSvHpcOz1XBbsbFtO+icyuI8tR05wqW3yekKtWXGn/dVG8O2QoNLecbqMkak3suhs3/E3XnOpRDiikAf44VIhNUGuwP4kZk9Y2b3XA6HhBBrw2o/xt/q7ifMbBOAH5vZq+7+xIW/0H4TuAcAhvp5lQ8hxNqyqju7u59o/z8B4GEAtwR+Z5+773X3vX3d6/BVfCEEgFUEu5n1mln/O48B/AGAly6XY0KIy8tqbrWjAB5ub/XnAfwvd//H2IR603BmMZzhM1kr03lP/Pyfg+O/sZtLLh95f1j6AYChSHHLJslsA4AcadOTy/GMpobztkURNQlHjh2htslFngHmPUPB8ayPSz+5oVlq6y4PUlu1wqWmKmmvNDDEX7OBPm6bOHWK2mbO84KT/cXwJd7VzWW+t85zcanQv4nazpx6i9r6TvM13jwQ9qXbIpmKpAgrIrLyJQe7ux8G8IFLnS+E6CyS3oRIBAW7EImgYBciERTsQiSCgl2IROhsr7eshPxguODgwjn+vlMrhgsKTi6EpTAAWKjy3mADRZ7Z1iR9t9rG4HCW8Yy9SpVLPGd48hrOznIJMFYQcWhjOJtrvjlD54yA+5hFMtGqBb6Olfmw1FSZ437sHN1AbQtEQgOACZLZBgBWCMuU05O8mCMiBUQX53lGXFbk18HEDM86HCfZcjtH+PWdYwlxsRaH3CSE+HVCwS5EIijYhUgEBbsQiaBgFyIROrob39Xdi/f91v+XBQsAOP4vr9F5fYPh3fhbPhw+FgD0ZMeorUp2igEgl+dJLVYI70w3vEzn9G/aTm3PHzhEbX1lvjO9def7qc1z4d3nQmTnvLkUbhkFANVqpMVWZK0yksTx8gsH6JyBUqRFUi9PkumN1LU7eSpcM65OlBUAyMgOPgAM9XN1YrrBk57OT3LbkVPTwfEto5vpnDxTlCLZVbqzC5EICnYhEkHBLkQiKNiFSAQFuxCJoGAXIhE6Kr3lsjx6BsOS0s6rr6PzFolqsWPXtXTOSI1LK1NHuCxXiyTCNOrhRIdbbvs4nbPjat4Ra9dvHqW2Z557gdqG+rgkc3IiXD8t77yMd6nAJS/wZcRcJClkmtSFG+rl54qcCo2IVDayMSzNAsBSLfx6nj0flrsAwCItu/ojdfLyGQ+naoUn3hx++3hwfGOZy3y7t4XbqHnk/q07uxCJoGAXIhEU7EIkgoJdiERQsAuRCAp2IRJhWenNzO4H8McAJtz9xvbYMIDvAbgKwFEAn3R3XmTrnWPlcshK4Qylk6cP0nl7fvuDwfHeQV7zK5s9QW2NeqRFTqTW2eG3w9lytw6F6+oBAHq2UVN/L5djuvI8k6s7Uuusq0gytiJ11bZuGaO2V958k9qKRV7nb2Y2vFZXbdtN51x3/Q3UNjnJL6++gTK1nTw1ERy3HK/vVh7iNf6mI7Xksohk191TprbF2fB1cIhcbwDQXQyfq1aPZClSy//jWwDueM/YfQAed/fdAB5v/yyEuIJZNtjb/dbf+w2JuwA80H78AICPX163hBCXm0v9m33U3cfbj0+h1dFVCHEFs+oNOnd3RL7paGb3mNl+M9s/Pc1rhgsh1pZLDfbTZjYGAO3/w7sgANx9n7vvdfe9g4MDl3g6IcRqudRgfxTA3e3HdwN45PK4I4RYK1YivX0XwO0ARszsOIAvAPgSgIfM7LMAjgH45EpOZpah0BW+u1cqvCDi0lI47a0QkaB6evmniN5IS6NSxrPe+vLhfk3f2vdNOudP/u291FaYP0VtxVIkeynHfdx19dbg+MTkSTqnMsez1zZvGqG2yRkuHS5Vw6/n1dfyTMVrruWZj9PPPUtt87Nz1DYzH/ax3uAS1eJiuB0TAJTLg9TWcC6VDZR5tl+9Gn49sxzvD3Z8PPxhukqy/IAVBLu7f5qYfn+5uUKIKwd9g06IRFCwC5EICnYhEkHBLkQiKNiFSISOFpyEGSwLSxALEfmnsrAYHC9EenLNnuNZXsi49FYAL0Q4Vg5nSr1xkPdsO3mc27DA5bBjx49S202beY+7rTvDxSi3TPBvNM8f4gU4h0tlausvc1nu8OGjwfGxLWFpEACmZvg3LGsRqez0Gd6rrukWHLdIcciFiPRmOX5dhc/UojdSqBLNcJZd0cLXPQBUz4VlW4+U7dSdXYhEULALkQgKdiESQcEuRCIo2IVIBAW7EInQWenNAZCeXZlzaWVsJNwfrqeLS28/PcALJQ5FivLtHubZSV2lsOxSzHOp5szEUWprLvHihTuu4UUss8jz7hkYCo6PjPLCl+cmedbYdCSzrRFRNzeS/mv5iFxaIdlfQDyba7HCs8PqxEk2DgCVJZ6BWa/z++OGkU3UZsavq6KFr5+SRfoOejjjsxApeqk7uxCJoGAXIhEU7EIkgoJdiERQsAuRCB3djTcDCvlwMslgH09OKfeHbdbku5UzzhMPzp7nKQsj/XxJeovhHdVGLlwjDwCOnjxKbaNDvJ7Zzmt5K6QKPx2eeibcRuvEON/57+8L7+ADQKHAWzy9fOgt7gi5jzQj95elyG783DxPCikP83ZNdZIIM36aFkRGbz9/XfIZTzTp6eE1EYusLRcA1MKJPI35KTpldFN/cDxf4G2tdGcXIhEU7EIkgoJdiERQsAuRCAp2IRJBwS5EIqyk/dP9AP4YwIS739ge+yKAPwdwpv1rn3f3x1ZywszCUsjmTeHaaS0niYwTSYAY28YTSfZH5LAp45KdZ+E6eYMjPKlicIAnQBS6wvIJAFwVkd76BsOJQQDwP+//dnB8IbJWM4uT1LawyGsDFiJXz+ah8POuTPJ6d/Mk0QgABgf46/Lqa29Q2+nTZ4LjM5GWUeUyf2IDvX3UljnXRAtVvo4ZqUW4sZcfb7ArHEf5yO17JXf2bwG4IzD+VXff0/63okAXQqwfywa7uz8BgL/1CyF+JVjN3+z3mtkBM7vfzPhXsIQQVwSXGuxfB3ANgD0AxgF8mf2imd1jZvvNbP/U1NQlnk4IsVouKdjd/bS7N9y9CeAbAGjXAnff5+573X1vuVy+RDeFEKvlkoLdzMYu+PETAF66PO4IIdaKlUhv3wVwO4ARMzsO4AsAbjezPWhVlTsK4C9WcrJcLkezfwaGuPRWb4TdLOV5JtF1u3ZQ2/5nuOQ1U7iW2po2Gxwf3crltVcO/gu1/c7v/Rm1/eLnfN78fKRNUvVscHzi1Nt0Tuw9f67GbXlwaWgoF86y29rNfZ8+wyW0esa3hUY3cVujEc6kW4y0eKos8rp785EaevUml/NqlRPUtqkQzujb0sez6Jbq4Tmxu/eywe7unw4Mf3O5eUKIKwt9g06IRFCwC5EICnYhEkHBLkQiKNiFSISOFpzM5XLo7QtnLw2NjNB5dQu7WckV6ZyuvgFqK5d5QcG33j5Fbbd+8P1hP+Z4O6me/nDWFQCMnzhObYdef53a6g3enihH6g3Oz0zTOf0bxqhteprLUIN9vBjl+667MTj+9Auv0jnPvnqU2m69/Q+prVDkEtXhQ4eC49Oz/HnFimJWFrm8tnOUS7rdvbyg6vBweJ7neQHOejVc+NJJVimgO7sQyaBgFyIRFOxCJIKCXYhEULALkQgKdiESoaPSm3sTzXpY8hgc5oX85hfDhQgXGrzvVpbx97Ed27dR2+sv88yr6YWwxNbXyzPstl9DTTj2Oi++eOLkOLV9+MMfpLaFhbA01L9lK50zvIUX53xrkktli0tcciz2hvuvDWzcTufc1M9flzNnwv3QAODosReobX4xLFNOTXMJbePGjdQ26Px12dnHJdFNA7wHW8HCmYDVGu9v10skthx4TOjOLkQiKNiFSAQFuxCJoGAXIhEU7EIkQkd345v1GmbPhXczuyO1vZYq4V1Oa3L3zfiu5Mgwb5/0eu4wtU1Mhlv4nMv4rvRgH6+td/2NPCHn8DFeM67GuyRhaiasduzevZvO2b2LSwbHxnkCzcsvv0ht586Gk1OKJa66DPXxRJLjL3NV4NQ5XtfOSLJUFmm9FWsdtpPnmWBHP08M6srxpJalSvj6aTZ5bcNanRyPX/a6swuRCgp2IRJBwS5EIijYhUgEBbsQiaBgFyIRVtL+aTuAvwMwitbG/j53/5qZDQP4HoCr0GoB9Ul3D/f8abO0tITDh8LS1o7dv0HndeXC0luzyhMF8l0RGSRi6+/n0lDfQLiu3fXXv4/O+cmPHqO2hWle765neBO1HTo+QW3bt4WTcna972Y6p1Tkl8HVO3iSz9Qkf7lfORhOKGo61w1PTPFEkhmSDAUAlQaXbWemwlLkps086eatc7w+3fB2LpeeK3E/0OTPbaoefm6e59fpEjleFTzhZiV39jqAv3H3GwB8CMBfmdkNAO4D8Li77wbwePtnIcQVyrLB7u7j7v5s+/EsgIMAtgK4C8AD7V97AMDH18hHIcRl4KL+ZjezqwDcBOBJAKPuv0zuPYXWx3whxBXKioPdzPoA/ADA59z9Xd9PdHcH+aKemd1jZvvNbP/sLC8YIIRYW1YU7GZWQCvQv+PuP2wPnzazsbZ9DEBw18jd97n7XnffG9v8EkKsLcsGu5kZWv3YD7r7Vy4wPQrg7vbjuwE8cvndE0JcLlaS9fa7AD4D4EUze7499nkAXwLwkJl9FsAxAJ9c7kALS3U8fygsG+248RY6r4lwtpmxzB8AaPL0n5nZWWqbmjpLbRuG9wTH77zjI3TOng9cT20P/fBhajPjEsrg4BC1bd0SlpT6Bsp0TlYPry8ADG/ml8jYrhq1TXeHZaPnXuD14sbneEqZF3g7r8HNPItx5JqwVJZFZK2Gcz9e83D7MgA4dIrLg8WMH3OxUgmOL0Qu73ozfH3MNnh24LLB7u4/A8A8/f3l5gshrgz0DTohEkHBLkQiKNiFSAQFuxCJoGAXIhE6WnCy0jC8Pt0dtJ1t8AKAXghLE7kqL4boRJoAgFyO27aM8Wyzf/074cyxrgKXXHbt5G2X/uhPP0Vt33/4H6jt7Cn+vMenw8ULK5VDdE4RXOOZXOS2Q8d41h6qYVnOR3iG4NCmcJFKAGhGKim2vvNF5nWFj9m0cCFKAKhF2opNN/i5ugr8mF15Lr3NWzjLrlbg5/JmeH0bEclWd3YhEkHBLkQiKNiFSAQFuxCJoGAXIhEU7EIkQkelt6WG4fWp8PvLIz/jfcP27BwJjm8u8gyknkIkW2sz7782NsKzq665mhQpdF5McPzMOWq7/0Eurz37/CvUxnrfAQBNBHT+vu4NfrxGia9HI8eloTzCEms9Ig3Vc+E5ANAVu1IjWWqVavh5e47PyUcy4rIm7+vnFS5T1sHnFZphHzPjr1m1FvY/0uJQd3YhUkHBLkQiKNiFSAQFuxCJoGAXIhE6uhvfgGEuF04WePzZ1+m8N94Mt4y647dvoHOu2cLb9Bw5HG5NBAC3ffBGausiiQmzVb7D/NA/Pk1tz71yktoW6pFWQpHd4lwh/P7djNTkyxnfRY7tWjeaPAFoieww1xp8jhmvabeESFKI8+eWz5Od7ozf53p6eEJLEdz/Bt9wR8N4qDXIxHqNvy7F/nJw3HL8PLqzC5EICnYhEkHBLkQiKNiFSAQFuxCJoGAXIhGWld7MbDuAv0OrJbMD2OfuXzOzLwL4cwBn2r/6eXd/LHqyfB4bRjYGbZPnuXwyfn4qOP7zF3irm0ZtZ8QTLq1s3EySXQBYFpbDntr/Ep3zDz/9BbUtNXnNNeS59JbLXfx7dGOJJ7t4RJZrRuS1mOTFWigV8vySs4xLmMj4a5aPzMuy8PliTUazyPrmnMuDjUiyUTMiHTLNbvNmLh/3D4Rtb5Yi68Q9+CV1AH/j7s+aWT+AZ8zsx23bV939v67gGEKIdWYlvd7GAYy3H8+a2UEAvGSqEOKK5KI+D5rZVQBuAvBke+heMztgZvebGW8tKoRYd1Yc7GbWB+AHAD7n7jMAvg7gGgB70Lrzf5nMu8fM9pvZ/voib5UshFhbVhTs1qrC/wMA33H3HwKAu59294a7NwF8A0Cwwbq773P3ve6+N9/NG0EIIdaWZYPdzAzANwEcdPevXDA+dsGvfQIA35IWQqw7K9mN/10AnwHwopk93x77PIBPm9ketOS4owD+YrkDmRmVSQoFLjXVK2E54ejpGTpnaf4gtd1283XU1l0eo7bpSlgi+ecn99M5FeeZS7U6l3FKJZ7Z1ozUQVtYCLcSipFFMrKMJ70h0pEJJSJ5xbKyELFZicuU3d28dl2eSH21SEbZ7Pw8tTUiMuVSnb8ug0PhOooAMDoWtvVFCu8tzob/JPbItbGS3fifAQi95FFNXQhxZaFv0AmRCAp2IRJBwS5EIijYhUgEBbsQidDRgpNwR7NOsqhiGUNZWIaqgmc7TcwtUduzr/FCj3cucGll1sNyx4nz/JuBpT6eXVVf4P5Xlrj/PT0RqYm0vYodz3Lcj1ykXVMsg82JjOaR+0shIjfO1Xj2XbXOpTImy8Uy9mIS2nyk9VZfmctr5Y285Vi1Hj7ma6/yrM4CyUasVbl/urMLkQgKdiESQcEuRCIo2IVIBAW7EImgYBciETosvQFgWUPO5Y4sCxfrazqXhRo5XuDv6ASXyu5/iOf3fPT2vcHxIyfPBMcBYKERK0IYkaG6eOHArMhtPaSHWbGby1qLs1y6imWHeUSiKpCMrSzPX7PYubJIUclYH7vFhbmLnhM7V3lomNo2jPKMybPnJqlt6uyp8PhbvCfhtbt2hQ0RSVF3diESQcEuRCIo2IVIBAW7EImgYBciERTsQiRCR6W3LJ9huFwO2ioVLofNL4YzeYoZz/6qR2ShXKS45RNPHaC2IyfD2XLT87xw5OTcIrWRZCcAQG9vJFsuUlSwVAo/t3xEruvq5hllWSQjLl/gx2yQ+0g9InlZxObOfWzU+PpXa+FF7u7iUuTIhg3UNjTC5bVqJHNzqRgpHkn6szXzXD6er4Svq2ZEwtadXYhEULALkQgKdiESQcEuRCIo2IVIhGV3482sC8ATAErt3/++u3/BzHYBeBDABgDPAPiMu0f2lwFvOpbILmIp8raz1AjvthYyvhtc55vI8Bw/Wa6b74IfIwkvuUhyR73Gd5hjikGlUqG2+Uh7ohx5bmyXHgB6i3zXtzuSQJPLcf+LXeHzdffw9a1WeSLM2UmeSNIEn5cvhNdjaKCXzhkdLlPb5s08EWZqntf5m506T21z01PB8fIwP9fZM2eD4/VIMtFK7uxLAD7q7h9Aqz3zHWb2IQB/C+Cr7n4tgPMAPruCYwkh1ollg91bvJMnWGj/cwAfBfD99vgDAD6+Fg4KIS4PK+3PnrU7uE4A+DGANwFMuf+yRelxAFvXxEMhxGVhRcHu7g133wNgG4BbAFy/0hOY2T1mtt/M9tcWeItlIcTaclG78e4+BeCfAHwYQNnsl429twE4Qebsc/e97r630DOwGl+FEKtg2WA3s41mVm4/7gbwMQAH0Qr6P23/2t0AHlkjH4UQl4GVJMKMAXjAzDK03hwecve/N7NXADxoZv8ZwHMAvrncgZrNJpYWw5JSKTM6r4d42azxJJNI1yI0wSWjWCJBk7SbqlcjCRwN/rxiLYhitmYkEYZJb+fPc+lnMrKOA31cohqM1GMbILXwusClvEaTS1d5iyTrlPiLvVQJH7OU569L7Fz1hemIjfs/N3WO2pokWaerxCXRCquTZ5HnRS1t3P0AgJsC44fR+vtdCPErgL5BJ0QiKNiFSAQFuxCJoGAXIhEU7EIkgsUknst+MrMzAI61fxwBEE7d6Szy493Ij3fzq+bHTnffGDJ0NNjfdWKz/e4ebp4mP+SH/LjsfuhjvBCJoGAXIhHWM9j3reO5L0R+vBv58W5+bfxYt7/ZhRCdRR/jhUiEdQl2M7vDzF4zs0Nmdt96+ND246iZvWhmz5vZ/g6e934zmzCzly4YGzazH5vZG+3/h9bJjy+a2Yn2mjxvZnd2wI/tZvZPZvaKmb1sZn/dHu/omkT86OiamFmXmT1lZi+0/fhP7fFdZvZkO26+Z2a84moId+/oPwAZWmWtrgZQBPACgBs67Ufbl6MARtbhvLcBuBnASxeM/RcA97Uf3wfgb9fJjy8C+PcdXo8xADe3H/cDeB3ADZ1ek4gfHV0TAAagr/24AOBJAB8C8BCAT7XH/zuAv7yY467Hnf0WAIfc/bC3Sk8/COCudfBj3XD3JwC8tzbyXWgV7gQ6VMCT+NFx3H3c3Z9tP55FqzjKVnR4TSJ+dBRvcdmLvK5HsG8F8PYFP69nsUoH8CMze8bM7lknH95h1N3H249PARhdR1/uNbMD7Y/5a/7nxIWY2VVo1U94Euu4Ju/xA+jwmqxFkdfUN+hudfebAfwhgL8ys9vW2yGg9c6O1hvRevB1ANeg1SNgHMCXO3ViM+sD8AMAn3P3d1Un7eSaBPzo+Jr4Koq8MtYj2E8A2H7Bz7RY5Vrj7ifa/08AeBjrW3nntJmNAUD7/4n1cMLdT7cvtCaAb6BDa2JmBbQC7Dvu/sP2cMfXJOTHeq1J+9xTuMgir4z1CPanAexu7ywWAXwKwKOddsLMes2s/53HAP4AwEvxWWvKo2gV7gTWsYDnO8HV5hPowJqYmaFVw/Cgu3/lAlNH14T50ek1WbMir53aYXzPbuOdaO10vgngP6yTD1ejpQS8AODlTvoB4LtofRysofW312fR6pn3OIA3APwEwPA6+fFtAC8COIBWsI11wI9b0fqIfgDA8+1/d3Z6TSJ+dHRNAPwWWkVcD6D1xvIfL7hmnwJwCMD/BlC6mOPqG3RCJELqG3RCJIOCXYhEULALkQgKdiESQcEuRCIo2IVIBAW7EImgYBciEf4vt7E0CllzrOkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13a84f6a0>"
      ]
     },
     "execution_count": 214,
     "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": [
    "plt.imshow(tfm(img))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "color must be int or single-element tuple",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m/var/folders/8w/r6kg1v9x7bbfzf9dw9gjslc80000gn/T/ipykernel_7424/3888883200.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0maug\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtoImg\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg_tensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/var/folders/8w/r6kg1v9x7bbfzf9dw9gjslc80000gn/T/ipykernel_7424/1910888612.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m     14\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     15\u001b[0m         \u001b[0mx1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtfm_a\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 16\u001b[0;31m         \u001b[0mx2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtfm_b\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     17\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/data/auto_augment.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, img)\u001b[0m\n\u001b[1;32m    650\u001b[0m             self.ops, self.num_layers, replace=self.choice_weights is None, p=self.choice_weights)\n\u001b[1;32m    651\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mop\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mops\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 652\u001b[0;31m             \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    653\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    654\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/data/auto_augment.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, img)\u001b[0m\n\u001b[1;32m    351\u001b[0m         \u001b[0mmagnitude\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0.\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmagnitude\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mupper_bound\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    352\u001b[0m         \u001b[0mlevel_args\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlevel_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmagnitude\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhparams\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlevel_fn\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 353\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maug_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mlevel_args\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    354\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    355\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__repr__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/timm/data/auto_augment.py\u001b[0m in \u001b[0;36mtranslate_y_rel\u001b[0;34m(img, pct, **kwargs)\u001b[0m\n\u001b[1;32m     73\u001b[0m     \u001b[0mpixels\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpct\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     74\u001b[0m     \u001b[0m_check_args_tf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 75\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mImage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mAFFINE\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpixels\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     76\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     77\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/PIL/Image.py\u001b[0m in \u001b[0;36mtransform\u001b[0;34m(self, size, method, data, resample, fill, fillcolor)\u001b[0m\n\u001b[1;32m   2486\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"missing method data\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2487\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2488\u001b[0;31m         \u001b[0mim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnew\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfillcolor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2489\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmode\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"P\"\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpalette\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2490\u001b[0m             \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpalette\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpalette\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/3.8.11/envs/dl/lib/python3.8/site-packages/PIL/Image.py\u001b[0m in \u001b[0;36mnew\u001b[0;34m(mode, size, color)\u001b[0m\n\u001b[1;32m   2703\u001b[0m         \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpalette\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mImagePalette\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mImagePalette\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2704\u001b[0m         \u001b[0mcolor\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpalette\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetcolor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcolor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2705\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_new\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcore\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfill\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcolor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2706\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2707\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mTypeError\u001b[0m: color must be int or single-element tuple"
     ]
    }
   ],
   "source": [
    "aug(toImg(img_tensor))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "RandomResizedCropAndInterpolation??"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandAugment(n=2, ops=\n",
       "\tAugmentOp(name=AutoContrast, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Equalize, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Invert, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Rotate, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Posterize, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Solarize, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=SolarizeAdd, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Color, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Contrast, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Brightness, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=Sharpness, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=ShearX, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=ShearY, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=TranslateXRel, p=0.5, m=9, mstd=0.5)\n",
       "\tAugmentOp(name=TranslateYRel, p=0.5, m=9, mstd=0.5))"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ToPILImage(mode=tensor([[[-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.2842,\n",
       "           0.7468,  1.7778,  1.7650, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,  0.0976,  2.2869,\n",
       "           2.7833,  2.7960,  0.7213, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.0296,  1.7778,  2.8088,\n",
       "           2.2869,  0.0976, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242,  2.0323,  2.7833,  2.2869,\n",
       "          -0.2969, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242,  1.6505,  2.8088,  2.5415,  0.0976,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.1569,  2.4142,  2.7960,  0.9886, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242,  1.7778,  2.7960,  1.3832, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242,  2.7960,  2.7833,  0.8741, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242,  1.1414,  2.8088,  2.2869, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242,  0.2249,  0.7468,  1.5105,  0.4795,\n",
       "          -0.1569, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           0.0976,  2.6687,  2.7960,  0.2122, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.1569,  1.9051,  2.7960,  2.7833,  2.7960,  2.7833,\n",
       "           2.4142,  0.0849, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           0.2377,  2.7960,  2.4269, -0.1696, -0.4242, -0.4242, -0.2842,\n",
       "           1.2559,  2.8088,  2.7960,  2.8088,  2.2869,  2.1596,  1.1286,\n",
       "           2.8088,  2.2869, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           1.2559,  2.7833,  1.1286, -0.4242, -0.4242,  0.0976,  1.7778,\n",
       "           2.7833,  2.0196,  0.4668,  0.2122, -0.2969, -0.4242, -0.4242,\n",
       "           2.2869,  2.7833, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           2.2996,  2.7960,  0.0976, -0.4242, -0.0296,  1.7778,  2.6815,\n",
       "           1.6378, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           1.5232,  2.7960, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           2.7960,  2.7833, -0.4242, -0.4242,  1.5105,  2.7833,  1.1286,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           2.0323,  2.7833, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           2.0323,  2.7960,  0.6195, -0.4242,  2.2996,  1.4978, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,  0.3649,\n",
       "           2.8088,  1.7650, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           1.0013,  2.7833,  2.4142, -0.1696,  2.7960,  1.4978, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.1569,  2.1596,\n",
       "           2.7960,  0.2122, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "           0.0976,  2.6687,  2.8215,  2.2869,  1.5105,  1.7650, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242,  0.3649,  2.2996,  2.7960,\n",
       "           1.1414, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242,  0.6068,  2.7960,  2.7833,  2.5415,  2.7833,  1.3832,\n",
       "           0.8741,  0.8741,  0.8741,  1.6505,  2.4142,  2.7960,  1.2305,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242,  0.6195,  2.4142,  2.8088,  2.7960,  2.8088,\n",
       "           2.7960,  2.8088,  2.7960,  2.8088,  2.5415,  0.6195, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.1696,  1.0013,  1.4978,  2.0196,\n",
       "           1.4978,  1.4978,  1.4978,  0.7341, -0.0424, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242],\n",
       "         [-0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242,\n",
       "          -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242, -0.4242]]]))"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "t = transforms.ToPILImage()\n",
    "t(example_data[0,:]).show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(6):\n",
    "    plt.subplot(2, 3, i+1)\n",
    "    plt.imshow(tfm(t(example_data[0,:])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "from timm.data.transforms import RandomResizedCropAndInterpolation\n",
    "from PIL import Image\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "tfm = RandomResizedCropAndInterpolation(size=224)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Linear(in_features=2048, out_features=1000, bias=True)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resnet.fc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# timm ResNet50\n",
    "\n",
    "\n",
    "import timm\n",
    "\n",
    "resnet = timm.create_model('resnet50', pretrained=True)\n",
    "\n",
    "class Encoder(nn.Module):\n",
    "    def __init__(self, model, embedding_dim=2048):\n",
    "        super().__init__()\n",
    "        self.model = model\n",
    "        self.model.fc = nn.Linear(2048, embedding_dim)\n",
    "    def forward(self, x):\n",
    "        x = self.model(x)\n",
    "        return x\n",
    "    \n",
    "class Projector(nn.Module):\n",
    "    def __init__(self, in_features, out_features):\n",
    "        super().__init__()\n",
    "        self.fc = nn.Sequential(\n",
    "            nn.Linear(in_features, out_features),\n",
    "            nn.BatchNorm1d(out_features),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(out_features, out_features),\n",
    "            nn.BatchNorm1d(out_features),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(out_features, out_features)\n",
    "        )\n",
    "    \n",
    "    def forward(self, x):\n",
    "        return self.fc(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv = nn.Conv2d(1, 3, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([4, 3, 28, 28])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conv(example_data).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([4, 1, 1, 28, 28])"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example_data[:, None, :].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoder = Encoder(resnet)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded = encoder(conv(example_data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([4, 2048])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoded.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "proj = Projector(2048, 8192)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding = proj(encoded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = example_data\n",
    "y_a, y_b = augment(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for x in loader:\n",
    "\n",
    "y_a, y_b = augment(x)\n",
    "\n",
    "z_a = (y_a)\n",
    "z_b = f(y_b)\n",
    "\n",
    "z_a_norm = (z_a - z_a.mean(0)) / z_a.std(0)\n",
    "z_b_norm = (z_b - z_b.mean(0)) / z_b.std(0)\n",
    "\n",
    "c = torch.mul(z_a_norm.T, z_b_norm)\n",
    "\n",
    "c_diff = (c - torch.eye(D)).pow(2)\n",
    "\n",
    "off_diagonal(c_diff).mul_(lambda_)\n",
    "loss = c_diff.sum()\n",
    "\n",
    "loss.backward()\n",
    "opt.step()\n"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.11"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}