cnn_clf.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "import os\n",
    "import time\n",
    "import csv\n",
    "\n",
    "os.environ['CUDA_DEVICE_ORDER'] = 'PCI_BUS_ID' \n",
    "os.environ['CUDA_VISIBLE_DEVICES']='0'\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "# from tqdm import tqdm\n",
    "import h5py\n",
    "# import imageio\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "# from torchvision import datasets, transforms\n",
    "# from torch.autograd import Variable\n",
    "import torch.autograd as autograd\n",
    "import torchvision.utils as vutils\n",
    "\n",
    "# from torchsummary import summary\n",
    "\n",
    "from torch.utils.data import DataLoader, TensorDataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data(size):\n",
    "    hfs = 48 # half full size 96/2=48\n",
    "    hs = size//2\n",
    "    start = hfs-hs\n",
    "    end = hfs+hs\n",
    "    \n",
    "    X = np.load('X_clean.npy')[:, start:end, start:end]/255\n",
    "    X = np.moveaxis(X, -1, 1).astype(np.float32)\n",
    "    y = np.load('Y_clean.npy').reshape(-1,1).astype(np.float32)\n",
    "    wsi = np.load('WSI_clean.npy')\n",
    "    \n",
    "#     X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.2, random_state=1)\n",
    "#     X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.25, random_state=1)\n",
    "    \n",
    "#     X_0 = torch.from_numpy(X[WSI==0])\n",
    "#     X_1 = torch.from_numpy(X[WSI==1])\n",
    "#     Y_0 = torch.from_numpy(Y[WSI==0])\n",
    "#     Y_1 = torch.from_numpy(Y[WSI==1])\n",
    "#     return X_0, Y_0, X_1, Y_1\n",
    "\n",
    "    return X, y, wsi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y, wsi = load_data(48)\n",
    "\n",
    "X = torch.from_numpy(X)\n",
    "y = torch.from_numpy(y)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.25, random_state=0)\n",
    "\n",
    "trainloader = DataLoader(TensorDataset(X_train, y_train), batch_size=128, shuffle=True, num_workers=0, pin_memory=True)\n",
    "valloader = DataLoader(TensorDataset(X_val, y_val), batch_size=128, shuffle=True, num_workers=0, pin_memory=True)\n",
    "testloader = DataLoader(TensorDataset(X_test, y_test), batch_size=128, shuffle=True, num_workers=0, pin_memory=True)\n",
    "# plt.hist(y_train_small.numpy(), bins=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_Clf(model, validation_loader):\n",
    "    model.eval()        \n",
    "    with torch.no_grad():\n",
    "        acc = .0\n",
    "        for i, data in enumerate(validation_loader):\n",
    "            X = data[0].cuda()\n",
    "            y = data[1].cuda()\n",
    "#             predicted = torch.argmax(model(X), dim=1)\n",
    "            predicted = torch.round(model(X))\n",
    "            acc+=(predicted == y).sum()/float(predicted.shape[0])       \n",
    "    model.train()\n",
    "    return (acc/(i+1)).detach().item()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Classifier(nn.Module):\n",
    "    def __init__(self, size):\n",
    "        super(Classifier, self).__init__()\n",
    "        \n",
    "        s = size // 4\n",
    "        s*=s\n",
    "        n_filters = 16\n",
    "        \n",
    "        self.net = nn.Sequential(\n",
    "            nn.Conv2d(3, n_filters, 3, stride=1, padding=1),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.MaxPool2d(2, 2),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "            \n",
    "            nn.Conv2d(n_filters, 2*n_filters, 3, stride=1, padding=1),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.MaxPool2d(2, 2),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "            \n",
    "            nn.Conv2d(2*n_filters, 4*n_filters, 3, stride=1, padding=1),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "            nn.Flatten(),\n",
    "            \n",
    "            nn.Linear(4*n_filters*s, 256, bias=True),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "\n",
    "            nn.Linear(256, 128, bias=True),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "            \n",
    "            nn.Linear(128, 256, bias=True),\n",
    "            nn.LeakyReLU(inplace=True),\n",
    "            nn.Dropout2d(p=0.2),\n",
    "\n",
    "            nn.Linear(256, 1, bias=True),\n",
    "            nn.Sigmoid()\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.net(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Training:\n",
    "    def __init__(self, model, model_name, cuda_device='cuda:0', lr=1e-4, early_stopping_count=20):\n",
    "        self.model_name = model_name\n",
    "        self.early_stopping_count = early_stopping_count\n",
    "        \n",
    "        device = 'cpu'\n",
    "        if torch.cuda.is_available():\n",
    "            device = cuda_device\n",
    "        self.device = torch.device(device)\n",
    "        \n",
    "        self.Clf = model.to(self.device)\n",
    "\n",
    "        self.Clf_opt = optim.Adam(self.Clf.parameters(), lr=lr)\n",
    "        self.Clf_criterion = torch.nn.BCELoss()\n",
    "        \n",
    "        self.path = 'results/'+model_name\n",
    "        \n",
    "        self.best_acc = .0\n",
    "        self.best_acc_count = 0\n",
    "\n",
    "        if not os.path.exists(self.path):\n",
    "            os.mkdir(self.path)\n",
    "\n",
    "    def train(self, train_loader, validation_loader, epochs):\n",
    "\n",
    "        self.stats = {\n",
    "            'clf_loss': [],\n",
    "            'clf_acc': [],\n",
    "            'clf_loss_val': [],\n",
    "            'clf_acc_val': [],\n",
    "            \n",
    "        }\n",
    "\n",
    "        for epoch in range(epochs):\n",
    "            clf_loss = []\n",
    "            clf_acc = []\n",
    "\n",
    "            for i, data in enumerate(train_loader):\n",
    "                X = data[0].to(self.device)\n",
    "                y = data[1].to(self.device)\n",
    "                    \n",
    "                loss, acc = self._train_Clf(X, y)\n",
    "                clf_loss.append(loss)\n",
    "                clf_acc.append(acc)                                \n",
    "            \n",
    "#             if epoch % 10 == 0:\n",
    "            acc_val = eval_Clf(self.Clf, validation_loader)\n",
    "            clf_loss_m = sum(clf_loss)/len(clf_loss)\n",
    "            clf_acc_m = sum(clf_acc)/len(clf_acc)     \n",
    "            self.stats['clf_loss'].append(clf_loss_m)\n",
    "            self.stats['clf_acc'].append(clf_acc_m)\n",
    "#             self.stats['clf_loss_val'].append(loss_val)\n",
    "            self.stats['clf_acc_val'].append(acc_val)\n",
    "\n",
    "            self.best_acc_count+=1\n",
    "            if acc_val > self.best_acc:\n",
    "                self.best_acc = acc_val\n",
    "                self.best_acc_count = 0\n",
    "                self.save_best()\n",
    "    \n",
    "            print(\"Epoch: %d, loss: %f, acc: %.3f, acc_val: %.3f\"%(epoch, clf_loss_m, clf_acc_m, acc_val))\n",
    "            \n",
    "            if self.best_acc_count >= self.early_stopping_count:\n",
    "                break\n",
    "        print(\"Finished. Best acc: %.3f\"%(self.best_acc))\n",
    "\n",
    "#     def _eval_Clf(self, validation_loader):\n",
    "#         self.Clf.eval()        \n",
    "#         with torch.no_grad():\n",
    "#             acc = .0\n",
    "#             for i, data in enumerate(validation_loader):\n",
    "#                 X = data[0].to(self.device)\n",
    "#                 y = data[1].to(self.device)\n",
    "\n",
    "#                 predicted = torch.argmax(self.Clf(X), dim=1)\n",
    "#                 acc+=(predicted == y).sum()/float(predicted.shape[0])       \n",
    "#         self.Clf.train()\n",
    "#         return (acc/(i+1)).detach().item()\n",
    "\n",
    "    def _train_Clf(self, data, labels):\n",
    "        self.Clf_opt.zero_grad()\n",
    "        \n",
    "        predicted = self.Clf(data)\n",
    "    \n",
    "        loss = self.Clf_criterion(predicted, labels)\n",
    "        \n",
    "        loss.backward()\n",
    "        self.Clf_opt.step()    \n",
    "      \n",
    "        acc = (np.round(predicted.detach().cpu()) == labels.detach().cpu()).sum()/float(predicted.shape[0])\n",
    "#         acc = (torch.argmax(predicted.detach().cpu(), dim=1) == labels.detach().cpu()).sum()/float(predicted.shape[0])\n",
    "\n",
    "        return loss.detach().item(), acc.item()\n",
    "\n",
    "    def save_best(self):\n",
    "        torch.save(self.Clf.state_dict(), self.path+'/best_model.pth')\n",
    "        np.save(self.path+'/acc_val.npy', t.stats['clf_acc_val'], allow_pickle=False)\n",
    "        np.save(self.path+'/acc.npy', t.stats['clf_acc'], allow_pickle=False)\n",
    "        np.save(self.path+'/loss.npy', t.stats['clf_loss'], allow_pickle=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0, loss: 0.565547, acc: 0.707, acc_val: 0.763\n",
      "Epoch: 1, loss: 0.492278, acc: 0.770, acc_val: 0.760\n",
      "Epoch: 2, loss: 0.473175, acc: 0.781, acc_val: 0.767\n",
      "Epoch: 3, loss: 0.462140, acc: 0.787, acc_val: 0.753\n",
      "Epoch: 4, loss: 0.453099, acc: 0.791, acc_val: 0.755\n",
      "Epoch: 5, loss: 0.443535, acc: 0.797, acc_val: 0.767\n",
      "Epoch: 6, loss: 0.436583, acc: 0.801, acc_val: 0.778\n",
      "Epoch: 7, loss: 0.429634, acc: 0.805, acc_val: 0.777\n",
      "Epoch: 8, loss: 0.421366, acc: 0.810, acc_val: 0.777\n",
      "Epoch: 9, loss: 0.414453, acc: 0.814, acc_val: 0.772\n",
      "Epoch: 10, loss: 0.409215, acc: 0.817, acc_val: 0.784\n",
      "Epoch: 11, loss: 0.402636, acc: 0.821, acc_val: 0.772\n",
      "Epoch: 12, loss: 0.395074, acc: 0.826, acc_val: 0.780\n",
      "Epoch: 13, loss: 0.389194, acc: 0.829, acc_val: 0.778\n",
      "Epoch: 14, loss: 0.381473, acc: 0.833, acc_val: 0.792\n",
      "Epoch: 15, loss: 0.373718, acc: 0.838, acc_val: 0.795\n",
      "Epoch: 16, loss: 0.365419, acc: 0.843, acc_val: 0.798\n",
      "Epoch: 17, loss: 0.359576, acc: 0.846, acc_val: 0.809\n",
      "Epoch: 18, loss: 0.351546, acc: 0.850, acc_val: 0.794\n",
      "Epoch: 19, loss: 0.340455, acc: 0.855, acc_val: 0.802\n",
      "Epoch: 20, loss: 0.331304, acc: 0.860, acc_val: 0.807\n",
      "Epoch: 21, loss: 0.324200, acc: 0.863, acc_val: 0.798\n",
      "Epoch: 22, loss: 0.313482, acc: 0.869, acc_val: 0.813\n",
      "Epoch: 23, loss: 0.305410, acc: 0.873, acc_val: 0.810\n",
      "Epoch: 24, loss: 0.295866, acc: 0.878, acc_val: 0.816\n",
      "Epoch: 25, loss: 0.287837, acc: 0.882, acc_val: 0.816\n",
      "Epoch: 26, loss: 0.278023, acc: 0.886, acc_val: 0.802\n",
      "Epoch: 27, loss: 0.269084, acc: 0.890, acc_val: 0.821\n",
      "Epoch: 28, loss: 0.261466, acc: 0.892, acc_val: 0.818\n",
      "Epoch: 29, loss: 0.252833, acc: 0.898, acc_val: 0.818\n",
      "Epoch: 30, loss: 0.245043, acc: 0.901, acc_val: 0.817\n",
      "Epoch: 31, loss: 0.237303, acc: 0.905, acc_val: 0.821\n",
      "Epoch: 32, loss: 0.231857, acc: 0.906, acc_val: 0.814\n",
      "Epoch: 33, loss: 0.222905, acc: 0.910, acc_val: 0.820\n",
      "Epoch: 34, loss: 0.216297, acc: 0.914, acc_val: 0.816\n",
      "Epoch: 35, loss: 0.211244, acc: 0.915, acc_val: 0.819\n",
      "Epoch: 36, loss: 0.205155, acc: 0.918, acc_val: 0.819\n",
      "Epoch: 37, loss: 0.199521, acc: 0.921, acc_val: 0.821\n",
      "Epoch: 38, loss: 0.195218, acc: 0.923, acc_val: 0.818\n",
      "Epoch: 39, loss: 0.188702, acc: 0.926, acc_val: 0.819\n",
      "Epoch: 40, loss: 0.185546, acc: 0.927, acc_val: 0.823\n",
      "Epoch: 41, loss: 0.180204, acc: 0.930, acc_val: 0.820\n",
      "Epoch: 42, loss: 0.174555, acc: 0.932, acc_val: 0.821\n",
      "Epoch: 43, loss: 0.171203, acc: 0.934, acc_val: 0.815\n",
      "Epoch: 44, loss: 0.165500, acc: 0.936, acc_val: 0.830\n",
      "Epoch: 45, loss: 0.163472, acc: 0.936, acc_val: 0.824\n",
      "Epoch: 46, loss: 0.159373, acc: 0.938, acc_val: 0.825\n",
      "Epoch: 47, loss: 0.154223, acc: 0.940, acc_val: 0.817\n",
      "Epoch: 48, loss: 0.150991, acc: 0.942, acc_val: 0.820\n",
      "Epoch: 49, loss: 0.148485, acc: 0.943, acc_val: 0.818\n",
      "Epoch: 50, loss: 0.144909, acc: 0.945, acc_val: 0.824\n",
      "Epoch: 51, loss: 0.142182, acc: 0.945, acc_val: 0.821\n",
      "Epoch: 52, loss: 0.141315, acc: 0.946, acc_val: 0.823\n",
      "Epoch: 53, loss: 0.136200, acc: 0.947, acc_val: 0.819\n",
      "Epoch: 54, loss: 0.134390, acc: 0.949, acc_val: 0.819\n",
      "Epoch: 55, loss: 0.130827, acc: 0.951, acc_val: 0.821\n",
      "Epoch: 56, loss: 0.129783, acc: 0.951, acc_val: 0.820\n",
      "Epoch: 57, loss: 0.125705, acc: 0.952, acc_val: 0.827\n",
      "Epoch: 58, loss: 0.124060, acc: 0.953, acc_val: 0.821\n",
      "Epoch: 59, loss: 0.123388, acc: 0.953, acc_val: 0.821\n",
      "Epoch: 60, loss: 0.119995, acc: 0.955, acc_val: 0.825\n",
      "Epoch: 61, loss: 0.117980, acc: 0.956, acc_val: 0.824\n",
      "Epoch: 62, loss: 0.116442, acc: 0.956, acc_val: 0.825\n",
      "Epoch: 63, loss: 0.114610, acc: 0.957, acc_val: 0.822\n",
      "Epoch: 64, loss: 0.112325, acc: 0.958, acc_val: 0.822\n",
      "Finished. Best acc: 0.830\n"
     ]
    }
   ],
   "source": [
    "t = Training(Classifier(X.shape[2]), 'cnn_clf_48')\n",
    "t.train(trainloader, valloader, 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.8669698238372803, 0.8646843433380127)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = Classifier(48)\n",
    "model.load_state_dict(torch.load('results/cnn_clf_1_48/best_model.pth'))\n",
    "model.cuda()\n",
    "\n",
    "eval_Clf(model, valloader), eval_Clf(model, testloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Classifier(48)\n",
    "model.load_state_dict(torch.load('results/cnn_clf_48/best_model.pth'))\n",
    "model.cuda();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_Clf(model, valloader), eval_Clf(model, testloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.8299145698547363, 0.8211551308631897, 0.8675994873046875)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(np.load('results/cnn_clf_48/acc_val.npy')[-1], \n",
    "np.load('results/cnn_clf_0_48/acc_val.npy')[-1],\n",
    "np.load('results/cnn_clf_1_48/acc_val.npy')[-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f6eb8ad0b20>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14, 6))\n",
    "plt.plot(np.load('results/cnn_clf_48/acc_val.npy'), label='acc')\n",
    "plt.plot(np.load('results/cnn_clf_0_48/acc_val.npy'), label='acc_0')\n",
    "plt.plot(np.load('results/cnn_clf_1_48/acc_val.npy'), label='acc_1')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f03b0a49460>]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14, 6))\n",
    "plt.plot(np.load('results/cnn_clf_1_48/acc.npy'), label='acc')\n",
    "plt.plot(np.load('results/cnn_clf_1_48/acc_val.npy'), label='acc_val')\n",
    "plt.legend()\n",
    "\n",
    "plt.figure(figsize=(14, 6))\n",
    "plt.plot(np.load('results/cnn_clf_1_48/loss.npy'), label='loss')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = 'results/'+mixed_clf.model_name\n",
    "with open(path+'/mixed_clf_fake_loss.csv', 'w', newline='') as myfile:\n",
    "    wr = csv.writer(myfile)\n",
    "    wr.writerow(mixed_clf.stats['clf_loss'])\n",
    "\n",
    "with open(path+'/mixed_clf_fake_acc_val.csv', 'w', newline='') as myfile:\n",
    "    wr = csv.writer(myfile)\n",
    "    wr.writerow(mixed_clf.stats['clf_acc_val'])\n",
    "\n",
    "with open(path+'/mixed_clf_fake_acc.csv', 'w', newline='') as myfile:\n",
    "    wr = csv.writer(myfile)\n",
    "    wr.writerow(mixed_clf.stats['clf_acc'])\n",
    "    \n",
    "# print(np.argmax(wgan_ct.stats['clf_acc_val']))\n",
    "# wgan_ct.stats['clf_acc_val'][3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fe249b05e10>"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(18,5))\n",
    "plt.title('Training classifier on mixed dataset')\n",
    "acc_max = np.max(mixed_clf.stats['clf_acc_val'])\n",
    "acc_max_e = np.argmax(mixed_clf.stats['clf_acc_val'])\n",
    "plt.axhline(y=acc_max, color='k', linestyle='-', alpha=0.2, linewidth=1)\n",
    "plt.axvline(x=acc_max_e, color='k', linestyle='-', alpha=0.2, linewidth=1)\n",
    "plt.plot(mixed_clf.stats['clf_acc'], label='Train acc')\n",
    "plt.plot(mixed_clf.stats['clf_acc_val'], label='Validation acc')\n",
    "plt.plot(mixed_clf.stats['clf_loss'], label='Train loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.legend()"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}