mnist_ce.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000 10000\n"
     ]
    }
   ],
   "source": [
    "import torch as T\n",
    "import torch.nn as nn\n",
    "\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torchvision import datasets, transforms\n",
    "\n",
    "# ds_tr = datasets.MNIST('./', train=True, download=True)\n",
    "ds_tr = datasets.MNIST('./', train=True)\n",
    "ds_ts = datasets.MNIST('./', train=False)\n",
    "\n",
    "print(len(ds_tr), len(ds_ts))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fa01389fba514e238866a7d01dfe1ae1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=60000), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c2c56d000db74593a2c6bdb112bb40ee",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=10000), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "torch.Size([50]) torch.Size([50, 28, 28]) torch.Size([50])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(1)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from tqdm import tqdm_notebook as tqdm\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "class MNIST(Dataset):\n",
    "    def __init__(self, ds):\n",
    "        super().__init__()\n",
    "        \n",
    "        self.dat = []\n",
    "        \n",
    "        for img, lbl in tqdm(ds):\n",
    "            img = np.asarray(img).astype(np.float32) / 255.0\n",
    "            \n",
    "            self.dat.append([img, lbl])\n",
    "    \n",
    "    def __len__(self):\n",
    "        return len(self.dat)\n",
    "    \n",
    "    def __getitem__(self, idx):\n",
    "        img, lbl = self.dat[idx]\n",
    "        \n",
    "        return idx, img, lbl\n",
    "\n",
    "ld_tr = DataLoader(MNIST(ds_tr), batch_size=50, shuffle=True)\n",
    "ld_ts = DataLoader(MNIST(ds_ts), batch_size=100)\n",
    "\n",
    "for idx, img, lbl in ld_tr:\n",
    "    print(idx.shape, img.shape, lbl.shape)\n",
    "    \n",
    "    plt.figure()\n",
    "    plt.imshow(ds_tr[idx[0]][0])\n",
    "    plt.show()\n",
    "    plt.close()\n",
    "    \n",
    "    print(lbl[0])\n",
    "    \n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model(\n",
      "  (fc): Sequential(\n",
      "    (0): Linear(in_features=784, out_features=100, bias=True)\n",
      "    (1): ReLU()\n",
      "    (2): Linear(in_features=100, out_features=100, bias=True)\n",
      "    (3): ReLU()\n",
      "    (4): Linear(in_features=100, out_features=10, bias=True)\n",
      "  )\n",
      ")\n",
      "torch.Size([50, 10])\n"
     ]
    }
   ],
   "source": [
    "class Model(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        \n",
    "        self.fc = nn.Sequential(nn.Linear(28**2, 10**2), nn.ReLU(), \n",
    "                                nn.Linear(10**2, 10**2), nn.ReLU(), \n",
    "                                nn.Linear(10**2, 10))\n",
    "    \n",
    "    def forward(self, img):\n",
    "        img = img.view((img.shape[0], -1))\n",
    "        \n",
    "        out = self.fc(img)\n",
    "        \n",
    "        return out\n",
    "\n",
    "model = Model().cuda()\n",
    "print(model)\n",
    "\n",
    "out = model(img.cuda())\n",
    "print(out.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "EPOCHS = 10\n",
    "LR = 0.002\n",
    "\n",
    "loss_ce = nn.CrossEntropyLoss().cuda()\n",
    "optim = T.optim.Adam(model.parameters(), lr=LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "73e2fdb2825946ce8ffdb9cf37f7ffa6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=10), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "394ece2dc30a435da2eede66132a740e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 1: 0.2641\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4a28b61bf66145a3a2edaf381e8d4576",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 95.9400%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b9fdd9ad0f3d43a597fe55cae6ef35b9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 2: 0.1072\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "db7de10a8c1f49f285a1a7603e7e1d3a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 96.8300%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bd19ac8af72b4021b61134a0982252f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 3: 0.0768\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "48efd25f26bb47298239d85bb8781b51",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.3800%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "05dd96d4d05f4957b73061250351d6d4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 4: 0.0604\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "714ab4132acc404d8fba3bfe24b5f12d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 96.9700%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b25c27d08754485daff4821bc72af434",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 5: 0.0518\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4d6c4d4f4d174981b4b85a8bc9418a78",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.4000%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e124b5e58e74b769c3350191e55202a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 6: 0.0412\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3ea276c956bb47b4b512a0943dc6bbc5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.5700%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "41af6d4d539c452cb180518586427d42",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 7: 0.0360\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "966ee5bb612a4d5595a661214e51680f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.5100%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e367568900ef4a9a9abcfc510532aacb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 8: 0.0317\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b3705e004124f63a5f3a27c293eddd7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.6700%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2609d4f5154a4f8e9df5fa3c32590a2f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 9: 0.0273\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ddd21e1cbe8b47ac9e9300fb21cfef76",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.4400%\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58962c57f7da4d9b8082928104aab491",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1200), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ep 10: 0.0262\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dcec8b61ec72456d880d553c84e79ba2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: 97.4900%\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for e in tqdm(range(EPOCHS)):\n",
    "    ls_ep = 0\n",
    "    \n",
    "    with tqdm(ld_tr) as TQ:\n",
    "        for _, img, lbl in TQ:\n",
    "            out = model(img.cuda())\n",
    "            \n",
    "            ls_bh = loss_ce(out, lbl.cuda())\n",
    "            \n",
    "            optim.zero_grad()\n",
    "            ls_bh.backward()\n",
    "            optim.step()\n",
    "            \n",
    "            ls_bh = ls_bh.detach().cpu().numpy()\n",
    "            ls_ep += ls_bh\n",
    "            \n",
    "            TQ.set_postfix(ls_bh='%.3f'%(ls_bh))\n",
    "        \n",
    "        ls_ep /= len(TQ)\n",
    "        print('Ep %d: %.4f' % (e+1, ls_ep))\n",
    "    \n",
    "    acc_ep = 0\n",
    "    with tqdm(ld_ts) as TQ:\n",
    "        for _, img, lbl in TQ:\n",
    "            out = model(img.cuda())\n",
    "            \n",
    "            out = np.argmax(out.detach().cpu().numpy(), axis=1)\n",
    "            lbl = lbl.numpy()\n",
    "            \n",
    "            acc_bh = np.average(out==lbl)\n",
    "            acc_ep += acc_bh\n",
    "            \n",
    "            TQ.set_postfix(acc_bh='%.3f%%'%(acc_bh*100))\n",
    "        \n",
    "        acc_ep /= len(TQ)\n",
    "        print('Test: %.4f%%'%(acc_ep*100))"
   ]
  }
 ],
 "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}