digits.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from utils.draw import draw_digits\n",
    "from torch.utils.data import DataLoader\n",
    "from torch.optim.lr_scheduler import StepLR\n",
    "from torchvision import datasets, transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNN(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(CNN, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)\n",
    "        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)\n",
    "        self.conv2_drop = nn.Dropout2d()\n",
    "        self.fc1 = nn.Linear(320, 50)\n",
    "        self.fc2 = nn.Linear(50, 10)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = x.view(-1, 1, 28, 28)\n",
    "        x = F.relu(F.max_pool2d(self.conv1(x), 2))\n",
    "        x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))\n",
    "        x = x.view(-1, 320)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = F.dropout(x, training=self.training)\n",
    "        x = self.fc2(x)\n",
    "        return F.softmax(x, dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "digits = datasets.MNIST('data', download=True,\n",
    "                        transform=transforms.Compose([\n",
    "                            transforms.ToTensor(),\n",
    "                            transforms.Lambda(lambda x: x.reshape(28*28))\n",
    "                        ]),\n",
    "                        target_transform=transforms.Compose([\n",
    "                            transforms.Lambda(lambda y: \n",
    "                                              torch.zeros(10, dtype=torch.float).scatter_(0, torch.tensor(y), value=1))\n",
    "                        ])\n",
    "                     )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x504 with 120 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_digits(digits)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "# Use the nn package to define our model and loss function.\n",
    "model = CNN()\n",
    "model = model.to(device)\n",
    "\n",
    "cost = torch.nn.BCELoss()\n",
    "\n",
    "# optimizer which Tensors it should update.\n",
    "learning_rate = 1e-3\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)\n",
    "scheduler = StepLR(optimizer, 3)\n",
    "\n",
    "# dataset!\n",
    "dataloader = DataLoader(digits, batch_size=64, num_workers=0, pin_memory=True)\n",
    "\n",
    "epochs = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "epoch 0, (lr: 1.0e-03)\n",
      "-------------------------------\n",
      "loss:   0.325435  [    0/60000]\n",
      "loss:   0.141028  [ 6400/60000]\n",
      "loss:   0.092830  [12800/60000]\n",
      "loss:   0.084161  [19200/60000]\n",
      "loss:   0.069875  [25600/60000]\n",
      "loss:   0.077485  [32000/60000]\n",
      "loss:   0.045714  [38400/60000]\n",
      "loss:   0.071020  [44800/60000]\n",
      "loss:   0.078321  [51200/60000]\n",
      "loss:   0.096473  [57600/60000]\n",
      "\n",
      "epoch 1, (lr: 1.0e-03)\n",
      "-------------------------------\n",
      "loss:   0.044585  [    0/60000]\n",
      "loss:   0.033604  [ 6400/60000]\n",
      "loss:   0.036741  [12800/60000]\n",
      "loss:   0.258524  [19200/60000]\n",
      "loss:   0.044610  [25600/60000]\n",
      "loss:   0.054129  [32000/60000]\n",
      "loss:   0.035628  [38400/60000]\n",
      "loss:   0.049562  [44800/60000]\n",
      "loss:   0.100081  [51200/60000]\n",
      "loss:   0.093859  [57600/60000]\n",
      "\n",
      "epoch 2, (lr: 1.0e-03)\n",
      "-------------------------------\n",
      "loss:   0.033547  [    0/60000]\n",
      "loss:   0.027782  [ 6400/60000]\n",
      "loss:   0.052631  [12800/60000]\n",
      "loss:   0.041481  [19200/60000]\n",
      "loss:   0.019959  [25600/60000]\n",
      "loss:   0.043494  [32000/60000]\n",
      "loss:   0.031453  [38400/60000]\n",
      "loss:   0.035974  [44800/60000]\n",
      "loss:   0.072765  [51200/60000]\n",
      "loss:   0.096724  [57600/60000]\n",
      "\n",
      "epoch 3, (lr: 1.0e-04)\n",
      "-------------------------------\n",
      "loss:   0.028677  [    0/60000]\n",
      "loss:   0.028074  [ 6400/60000]\n",
      "loss:   0.016807  [12800/60000]\n",
      "loss:   0.057601  [19200/60000]\n",
      "loss:   0.018693  [25600/60000]\n",
      "loss:   0.061979  [32000/60000]\n",
      "loss:   0.035869  [38400/60000]\n",
      "loss:   0.059403  [44800/60000]\n",
      "loss:   0.055504  [51200/60000]\n",
      "loss:   0.043458  [57600/60000]\n",
      "\n",
      "epoch 4, (lr: 1.0e-04)\n",
      "-------------------------------\n",
      "loss:   0.023257  [    0/60000]\n",
      "loss:   0.011421  [ 6400/60000]\n",
      "loss:   0.029147  [12800/60000]\n",
      "loss:   0.053576  [19200/60000]\n",
      "loss:   0.028429  [25600/60000]\n",
      "loss:   0.043614  [32000/60000]\n",
      "loss:   0.030894  [38400/60000]\n",
      "loss:   0.048809  [44800/60000]\n",
      "loss:   0.055497  [51200/60000]\n",
      "loss:   0.055366  [57600/60000]\n",
      "\n",
      "epoch 5, (lr: 1.0e-04)\n",
      "-------------------------------\n",
      "loss:   0.014803  [    0/60000]\n",
      "loss:   0.027200  [ 6400/60000]\n",
      "loss:   0.023097  [12800/60000]\n",
      "loss:   0.050086  [19200/60000]\n",
      "loss:   0.018372  [25600/60000]\n",
      "loss:   0.050498  [32000/60000]\n",
      "loss:   0.030511  [38400/60000]\n",
      "loss:   0.034882  [44800/60000]\n",
      "loss:   0.056668  [51200/60000]\n",
      "loss:   0.060304  [57600/60000]\n",
      "\n",
      "epoch 6, (lr: 1.0e-05)\n",
      "-------------------------------\n",
      "loss:   0.019734  [    0/60000]\n",
      "loss:   0.029140  [ 6400/60000]\n",
      "loss:   0.031578  [12800/60000]\n",
      "loss:   0.067004  [19200/60000]\n",
      "loss:   0.007449  [25600/60000]\n",
      "loss:   0.051854  [32000/60000]\n",
      "loss:   0.021375  [38400/60000]\n",
      "loss:   0.035755  [44800/60000]\n",
      "loss:   0.048635  [51200/60000]\n",
      "loss:   0.062883  [57600/60000]\n",
      "\n",
      "epoch 7, (lr: 1.0e-05)\n",
      "-------------------------------\n",
      "loss:   0.024772  [    0/60000]\n",
      "loss:   0.026985  [ 6400/60000]\n",
      "loss:   0.020415  [12800/60000]\n",
      "loss:   0.037523  [19200/60000]\n",
      "loss:   0.022117  [25600/60000]\n",
      "loss:   0.052355  [32000/60000]\n",
      "loss:   0.033635  [38400/60000]\n",
      "loss:   0.031943  [44800/60000]\n",
      "loss:   0.053595  [51200/60000]\n",
      "loss:   0.037792  [57600/60000]\n",
      "\n",
      "epoch 8, (lr: 1.0e-05)\n",
      "-------------------------------\n",
      "loss:   0.018667  [    0/60000]\n",
      "loss:   0.043463  [ 6400/60000]\n",
      "loss:   0.019595  [12800/60000]\n",
      "loss:   0.055454  [19200/60000]\n",
      "loss:   0.036523  [25600/60000]\n",
      "loss:   0.036935  [32000/60000]\n",
      "loss:   0.030965  [38400/60000]\n",
      "loss:   0.046444  [44800/60000]\n",
      "loss:   0.216413  [51200/60000]\n",
      "loss:   0.054808  [57600/60000]\n",
      "\n",
      "epoch 9, (lr: 1.0e-06)\n",
      "-------------------------------\n",
      "loss:   0.029028  [    0/60000]\n",
      "loss:   0.026590  [ 6400/60000]\n",
      "loss:   0.027678  [12800/60000]\n",
      "loss:   0.058059  [19200/60000]\n",
      "loss:   0.064346  [25600/60000]\n",
      "loss:   0.033335  [32000/60000]\n",
      "loss:   0.024782  [38400/60000]\n",
      "loss:   0.048361  [44800/60000]\n",
      "loss:   0.061041  [51200/60000]\n",
      "loss:   0.034552  [57600/60000]\n"
     ]
    }
   ],
   "source": [
    "for t in range(epochs):\n",
    "    print('\\nepoch {}, (lr: {:>.1e})'.format(t, scheduler.get_last_lr()[0]))\n",
    "    print('-------------------------------')\n",
    "    for batch, (X, Y) in enumerate(dataloader):\n",
    "        X, Y = X.to(device), Y.to(device)\n",
    "        optimizer.zero_grad()\n",
    "        pred = model(X)\n",
    "        loss = cost(pred, Y)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        if batch % 100 == 0:\n",
    "            print('loss: {:>10f}  [{:>5d}/{:>5d}]'.format(loss.item(), batch * len(X), len(dataloader.dataset)))\n",
    "            \n",
    "    # step scheduler\n",
    "    scheduler.step()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Does it work???"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data = datasets.MNIST('data', train=False, download=True,\n",
    "                        transform=transforms.Compose([\n",
    "                            transforms.ToTensor(),\n",
    "                            transforms.Lambda(lambda x: x.reshape(28*28))\n",
    "                        ]),\n",
    "                        target_transform=transforms.Compose([\n",
    "                            transforms.Lambda(lambda y: \n",
    "                                              torch.zeros(10, dtype=torch.float).scatter_(0, torch.tensor(y), value=1))\n",
    "                        ])\n",
    "                     )\n",
    "test_loader = DataLoader(digits, batch_size=64, num_workers=0, pin_memory=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Error:\n",
      "acc: 97.8%, avg loss: 0.000202\n"
     ]
    }
   ],
   "source": [
    "model.eval()\n",
    "test_loss = 0\n",
    "correct = 0\n",
    "with torch.no_grad():\n",
    "    for batch, (X, Y) in enumerate(test_loader):\n",
    "        X, Y = X.to(device), Y.to(device)\n",
    "        pred = model(X)\n",
    "\n",
    "        test_loss += cost(pred, Y).item()\n",
    "        correct += (pred.argmax(1) == Y.argmax(1)).type(torch.float).sum().item()\n",
    "\n",
    "test_loss /= len(dataloader.dataset)\n",
    "correct /= len(dataloader.dataset)\n",
    "print('Test Error:')\n",
    "print('acc: {:>0.1f}%, avg loss: {:>8f}'.format(100*correct, test_loss))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Saving Things!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saved onnx model to \"superfile.onnx\"\n"
     ]
    }
   ],
   "source": [
    "import torch.onnx as onnx\n",
    "\n",
    "# create dummy variable to traverse graph\n",
    "x = torch.randint(255, (1, 28*28), dtype=torch.float).to(device) / 255\n",
    "onnx.export(model, x, 'superfile.onnx')\n",
    "print('Saved onnx model to \"superfile.onnx\"')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "file_extension": ".py",
  "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"
  },
  "mimetype": "text/x-python",
  "name": "python",
  "npconvert_exporter": "python",
  "pygments_lexer": "ipython3",
  "version": 3
 },
 "nbformat": 4,
 "nbformat_minor": 2
}