EfficientNetB3_V1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Diabetic Retinopathy\n",
    "# Dataset - https://www.kaggle.com/competitions/diabetic-retinopathy-detection/data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "from sklearn.model_selection import train_test_split as datasetSplit\n",
    "import numpy as np \n",
    "import os\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchvision \n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "\n",
    "from PIL import Image\n",
    "import albumentations as A\n",
    "from albumentations.pytorch import ToTensorV2\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.metrics import cohen_kappa_score\n",
    "from efficientnet_pytorch import EfficientNet\n",
    "from tqdm import tqdm\n",
    "\n",
    "from torch.optim import lr_scheduler\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataset paths containing all the images and image labels\n",
    "dataset_images = \"Data/images_650/\"\n",
    "dataset_labels = \"Data/trainLabels.csv\"\n",
    "\n",
    "BATCHSIZE = 64\n",
    "DEVICE = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "LEARNING_RATE = 0.0001\n",
    "\n",
    "WEIGHT_DECAY = 0.0005 #L2 Norm on the model weights\n",
    "NUM_EPOCHS = 100\n",
    "NUM_WORKERS = 0 \n",
    "CHECKPOINT_FOLDER = \"Checkpoints/\"\n",
    "CHECKPOINT_FILE = \"EfNetB3_v1.pth.tar\"\n",
    "PATIENCE = 3 #for early stopping\n",
    "\n",
    "PIN_MEMORY = True\n",
    "SAVE_MODEL = True\n",
    "LOAD_MODEL = True\n",
    "\n",
    "INPUT_IMAGE_WIDTH = 120\n",
    "INPUT_IMAGE_HEIGHT = 120\n",
    "\n",
    "from torch.utils.tensorboard import SummaryWriter\n",
    "writer = SummaryWriter(\"runs/EfNetB3_v1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35126 images found in the dataset_images\n"
     ]
    }
   ],
   "source": [
    "print(f'{len(os.listdir(dataset_images))} images found in the dataset_images') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train Val Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "35126 images and 35126 targets found\n"
     ]
    }
   ],
   "source": [
    "inputs = np.array(pd.read_csv(dataset_labels)[\"image\"])\n",
    "targets = np.array(pd.read_csv(dataset_labels)[\"level\"])\n",
    "print(f\"{len(inputs)} images and {len(targets)} targets found\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_inputs, val_inputs, train_labels, val_labels = datasetSplit(inputs,targets,test_size= 0.11,random_state=47,shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset Train Size: 31262, Dataset Test Size:3864\n"
     ]
    }
   ],
   "source": [
    "print(f\"Dataset Train Size: {len(train_inputs)}, Dataset Test Size:{len(val_inputs)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dataset Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DiabeticRetinopathyDataset(Dataset):\n",
    "    def __init__(self,inputs,targets,data_transforms=None):\n",
    "        self.inputs = inputs\n",
    "        self.targets = torch.from_numpy(targets.astype(np.float32))\n",
    "        self.data_transforms = data_transforms\n",
    "        self.n_samples = len(inputs)\n",
    "        \n",
    "    def __getitem__(self,index):\n",
    "        img_path = self.inputs[index]\n",
    "        img = Image.open(dataset_images+img_path+\".jpeg\")\n",
    "        img = np.array(img) #Albumenations takes nd.array as input\n",
    "        # img = torch.from_numpy(img.astype(np.float32))\n",
    "        img_label = self.targets[index]\n",
    "        if self.data_transforms:\n",
    "            img = self.data_transforms(image=img)[\"image\"]\n",
    "            img.type(torch.float32)\n",
    "        return img, img_label, img_path\n",
    "        \n",
    "    def __len__(self):\n",
    "        return self.n_samples\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data Transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_transforms = A.Compose([\n",
    "    A.Resize(width=INPUT_IMAGE_WIDTH+30,height=INPUT_IMAGE_HEIGHT+30),\n",
    "    A.RandomCrop(width=INPUT_IMAGE_WIDTH,height=INPUT_IMAGE_HEIGHT),\n",
    "    A.Normalize(mean=[0.3199, 0.2240, 0.1609],std=[0.3020, 0.2183, 0.1741],max_pixel_value = 255.0), \n",
    "    ToTensorV2()\n",
    "    ])\n",
    "\n",
    "#Test with more transformations\n",
    "\n",
    "val_transforms = A.Compose([\n",
    "    A.Resize(width=INPUT_IMAGE_WIDTH,height=INPUT_IMAGE_HEIGHT),\n",
    "    A.Normalize(mean=[0.3199, 0.2240, 0.1609],std=[0.3020, 0.2183, 0.1741],max_pixel_value = 255.0), \n",
    "    ToTensorV2()\n",
    "    ])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dataset Loader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = DiabeticRetinopathyDataset(train_inputs,train_labels,data_transforms=train_transforms)\n",
    "val_dataset = DiabeticRetinopathyDataset(val_inputs,val_labels,data_transforms=val_transforms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader =DataLoader(train_dataset,batch_size=BATCHSIZE,shuffle=True,num_workers=NUM_WORKERS)\n",
    "val_loader =DataLoader(val_dataset,batch_size=BATCHSIZE,shuffle=False,num_workers=NUM_WORKERS)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Having a look at first batch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_iter = iter(train_loader)\n",
    "first_batch = train_iter.next()\n",
    "batch_images, batch_labels, batch_files = first_batch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([64, 3, 120, 120])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([64])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_labels.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Visualise Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2d7f7fdd190>"
      ]
     },
     "execution_count": 15,
     "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": [
    "batch_images_grid = torchvision.utils.make_grid(batch_images,nrow=8)\n",
    "writer.add_image(\"Retina Images\",batch_images_grid)\n",
    "writer.close()\n",
    "\n",
    "plt.imshow(batch_images_grid.permute(1, 2, 0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Utility Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_prediction(model,dataloader,output_csv=\"submission.csv\"):\n",
    "    all_preds = []\n",
    "    file_names = []\n",
    "    # total_loss = 0.0\n",
    "    model.eval() #set the model to evaluation state\n",
    "\n",
    "    for inputs, labels, names in enumerate(dataloader):\n",
    "        inputs = inputs.to(DEVICE)\n",
    "        with torch.no_grad():\n",
    "            outs = model(inputs)\n",
    "            _, batch_preds = torch.max(outs,1)\n",
    "            all_preds.append(batch_preds.cpu().numpy())\n",
    "            # batch_loss = nn.CrossEntropyLoss(outs,labels)\n",
    "            # total_loss += batch_loss\n",
    "            # batch_kappa = \n",
    "            file_names.append(names)\n",
    "\n",
    "    # epoch_loss = total_loss/len(dataloader)\n",
    "    df = pd.DataFrame({\"image\":file_names,\"level\":np.concatenate(all_preds, axis=0)})\n",
    "    df.to_csv(output_csv,index=False)\n",
    "    model.train() #Set the model back to training phase\n",
    "\n",
    "def calculate_accuracy(model,dataloader,epoch_num):\n",
    "    model.eval()\n",
    "    total_loss = 0.0\n",
    "    all_preds, all_labels = [],[]\n",
    "    correct_predictions = 0\n",
    "    num_samples = 0\n",
    "    dataiter = tqdm(dataloader)\n",
    "    for batchID, (inputs, labels, names) in enumerate(dataiter):\n",
    "        inputs = inputs.to(DEVICE)\n",
    "        labels = labels.type(torch.LongTensor)\n",
    "        labels = labels.to(DEVICE)\n",
    "\n",
    "        with torch.no_grad():\n",
    "            outs = model(inputs)\n",
    "            _, batch_preds = torch.max(outs,1)\n",
    "            batch_loss = loss_func(outs,labels)\n",
    "            correct_predictions += (batch_preds==labels).sum().item()\n",
    "            num_samples += outs.shape[0]\n",
    "\n",
    "            all_preds.append(batch_preds.cpu().numpy())\n",
    "            all_labels.append(labels.cpu().numpy())\n",
    "            total_loss += batch_loss.item()\n",
    "        dataiter.set_postfix(loss=batch_loss.item())\n",
    "            \n",
    "            \n",
    "\n",
    "    model.train()\n",
    "    epoch_acc = 100.0*correct_predictions/num_samples\n",
    "    epoch_loss = total_loss/len(dataloader)\n",
    "    # print(f\"Epoch Accuracy: {accuracy:.2f}%. Predicted {correct_predictions} correcctly out of {num_samples} images!\")\n",
    "    all_preds = np.concatenate(all_preds,axis=0,dtype=np.int64)\n",
    "    all_labels = np.concatenate(all_labels,axis=0,dtype=np.int64)\n",
    "    kappa = cohen_kappa_score(all_labels,all_preds,weights='quadratic')\n",
    "    print(f\"Epoch (Val): {epoch_num}, Loss= {epoch_loss}, Accuracy= {epoch_acc}, Kappa= {kappa}\")\n",
    "    writer.add_scalar('val_loss',epoch_loss,global_step=epoch_num)\n",
    "    writer.add_scalar('val_acc',epoch_acc,global_step=epoch_num)\n",
    "    writer.add_scalar('val_kappa',kappa,global_step=epoch_num)\n",
    "    writer.close()\n",
    "    return epoch_acc, kappa \n",
    "\n",
    "            \n",
    "def save_checkpoint(state_dicts):\n",
    "    torch.save(state_dicts,CHECKPOINT_FOLDER+CHECKPOINT_FILE)\n",
    "\n",
    "def load_checkpoint(checkpoint_file,model,optimiser,learning_rate):\n",
    "    checkpoint = torch.load(checkpoint_file)\n",
    "    model.load_state_dict(checkpoint[\"state_dict\"])\n",
    "    optimiser.load_state_dict(checkpoint[\"optimiser\"])\n",
    "\n",
    "    for params in optimiser.param_groups:\n",
    "        params[\"lr\"] = learning_rate\n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PreProcess\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Design the Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded pretrained weights for efficientnet-b3\n"
     ]
    }
   ],
   "source": [
    "model = EfficientNet.from_pretrained(\"efficientnet-b3\")\n",
    "in_features = model._fc.in_features\n",
    "model._fc = nn.Linear(in_features,5) #\n",
    "model = model.to(DEVICE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Loss Function, Optmiser"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_func = nn.CrossEntropyLoss()\n",
    "optimiser = torch.optim.Adam(model.parameters(),lr=LEARNING_RATE,weight_decay=WEIGHT_DECAY)\n",
    "scaler = torch.cuda.amp.GradScaler()\n",
    "lr_scheduler1 = lr_scheduler.StepLR(optimiser,step_size=7,gamma=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_epoch(model,dataloader,optimiser,scheduler,loss_func,scaler,epoch_num):\n",
    "    total_loss = 0.0\n",
    "    dataiter = tqdm(dataloader)\n",
    "    correct_predictions = 0\n",
    "    num_samples = 0\n",
    "    all_preds, all_labels = [],[]\n",
    "    for batch_num,(inputs,labels,filenames) in enumerate(dataiter):\n",
    "        inputs = inputs.to(DEVICE)\n",
    "        labels = labels.type(torch.LongTensor)\n",
    "        labels = labels.to(DEVICE)\n",
    "\n",
    "        if batch_num%40 == 0 and epoch_num%10==0 : #visualising few batch images (after transforms) \n",
    "            writer.add_image(\"Batch_\"+str(batch_num),torchvision.utils.make_grid(inputs,nrow=8))\n",
    "            writer.close()\n",
    "\n",
    "        #forward pass\n",
    "        with torch.cuda.amp.autocast():\n",
    "            outs = model(inputs)\n",
    "            _, batch_preds = torch.max(outs,1)\n",
    "            # print(outs.shape,labels.shape) \n",
    "            batch_loss = loss_func(outs,labels)\n",
    "\n",
    "        \n",
    "        total_loss += batch_loss.item()\n",
    "        correct_predictions += (batch_preds==labels).sum().item()\n",
    "        num_samples += outs.shape[0]\n",
    "        all_preds.append(batch_preds.cpu().numpy())\n",
    "        all_labels.append(labels.cpu().numpy())\n",
    "\n",
    "        #backward step\n",
    "        optimiser.zero_grad()\n",
    "        scaler.scale(batch_loss).backward()\n",
    "        scaler.step(optimiser)\n",
    "        # optimiser.step()\n",
    "        scaler.update()\n",
    "        dataiter.set_postfix(loss=batch_loss.item())\n",
    "\n",
    "    epoch_loss = total_loss/len(dataloader)\n",
    "    epoch_acc = 100.0*correct_predictions/num_samples\n",
    "    all_preds = np.concatenate(all_preds,axis=0,dtype=np.int64)\n",
    "    all_labels = np.concatenate(all_labels,axis=0,dtype=np.int64)\n",
    "    scheduler.step()\n",
    "\n",
    "    kappa = cohen_kappa_score(all_labels,all_preds,weights='quadratic')\n",
    "    print(f\"Epoch (Train): {epoch_num}, Loss= {epoch_loss}, Accuracy= {epoch_acc}, Kappa= {kappa}\")\n",
    "    writer.add_scalar('training_loss',epoch_loss,global_step=epoch_num)\n",
    "    writer.add_scalar('training_acc',epoch_acc,global_step=epoch_num)\n",
    "    writer.add_scalar('training_kappa',kappa,global_step=epoch_num)\n",
    "    writer.close()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Check if it is working for one epoch\n",
    "# train_epoch(model,train_loader,optimiser,loss_func,scaler,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def trainModel(model,trainloader,valloader,optimiser,lr_scheduler1,loss_func,epochs):\n",
    "    \n",
    "    checkpoint_stuff = {\"state_dict\":model.state_dict() ,\"optimiser\":optimiser.state_dict()}\n",
    "    best_kappa = 0.0\n",
    "    best_epoch = 0\n",
    "    es = 0 #early stopper\n",
    "\n",
    "    for epoch in range(epochs):\n",
    "\n",
    "        #Train Phasenb m, bn,\n",
    "        model.train()\n",
    "        \n",
    "        train_epoch(model,train_loader,optimiser,lr_scheduler1,loss_func,scaler,epoch)\n",
    "        \n",
    "        #Validation Phase\n",
    "        model.eval()\n",
    "        epoch_acc, kappa  = calculate_accuracy(model,val_loader,epoch)\n",
    "        \n",
    "        #Early stopping and checkpoints\n",
    "        if (kappa > best_kappa) :\n",
    "            best_kappa = kappa\n",
    "            es = 0 \n",
    "            checkpoint_stuff = {\"state_dict\":model.state_dict() ,\"optimiser\":optimiser.state_dict()}\n",
    "            best_epoch = epoch\n",
    "            print(f\"Checkpoint Saved at {best_epoch}\")\n",
    "        else :\n",
    "            es += 1\n",
    "            print(f\"Model performance has not improved in this epoch. Patience: {es}/{PATIENCE}\")\n",
    "            if es>PATIENCE:\n",
    "                print(f\"Early Stopping with {best_kappa} at Epoch {best_epoch}\")\n",
    "                break\n",
    "\n",
    "\n",
    "    model.load_state_dict(checkpoint_stuff[\"state_dict\"]) #load the best model\n",
    "    if SAVE_MODEL:\n",
    "        save_checkpoint(checkpoint_stuff)\n",
    "        # print(\"Checkpoint - Model Saved at {best_epoch}\")\n",
    "    return model #return the trained model\n",
    "    \n",
    "    \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:33<00:00,  1.79it/s, loss=0.608]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 0, Loss= 0.8586835028691282, Accuracy= 72.61851449043567, Kappa= 0.1472502113036287\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.30it/s, loss=0.902]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 0, Loss= 0.8327332158557704, Accuracy= 71.972049689441, Kappa= 0.4115324346711693\n",
      "Checkpoint Saved at 0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:25<00:00,  1.84it/s, loss=0.58] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 1, Loss= 0.7111397974330224, Accuracy= 75.92604439895081, Kappa= 0.4482600255410043\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.33it/s, loss=0.856]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 1, Loss= 0.7555856729140047, Accuracy= 74.71532091097309, Kappa= 0.502419168591224\n",
      "Checkpoint Saved at 1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:25<00:00,  1.85it/s, loss=0.53] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 2, Loss= 0.6625735596401316, Accuracy= 77.28872113108567, Kappa= 0.536436287049924\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.29it/s, loss=0.872]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 2, Loss= 0.7271277665114794, Accuracy= 76.08695652173913, Kappa= 0.516495539156226\n",
      "Checkpoint Saved at 2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:24<00:00,  1.85it/s, loss=0.419]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 3, Loss= 0.6206748989584012, Accuracy= 78.75375855671422, Kappa= 0.5949187212137448\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.34it/s, loss=0.976]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 3, Loss= 0.7375166582279518, Accuracy= 75.6728778467909, Kappa= 0.5175556879984915\n",
      "Checkpoint Saved at 3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:22<00:00,  1.86it/s, loss=0.47] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 4, Loss= 0.5866819740804426, Accuracy= 79.41590429275158, Kappa= 0.6262802505929218\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.34it/s, loss=1.04] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 4, Loss= 0.7546424597013192, Accuracy= 73.93892339544513, Kappa= 0.5329556802166246\n",
      "Checkpoint Saved at 4\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:22<00:00,  1.86it/s, loss=0.507]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 5, Loss= 0.5487087073379743, Accuracy= 80.58985349625743, Kappa= 0.6677621954865807\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.33it/s, loss=0.788]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 5, Loss= 0.7421661038867763, Accuracy= 74.5600414078675, Kappa= 0.5514317240486399\n",
      "Checkpoint Saved at 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:22<00:00,  1.86it/s, loss=0.597]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 6, Loss= 0.5077720077856918, Accuracy= 81.78299532979337, Kappa= 0.7038499808741079\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:25<00:00,  2.35it/s, loss=0.9]  \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 6, Loss= 0.7941715258066772, Accuracy= 72.17908902691511, Kappa= 0.5413728584279602\n",
      "Model performance has not improved in this epoch. Patience: 1/3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:22<00:00,  1.86it/s, loss=0.685]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 7, Loss= 0.41340084684772727, Accuracy= 85.00415840317318, Kappa= 0.7754062087162312\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.34it/s, loss=1.05] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 7, Loss= 0.7557862210469167, Accuracy= 75.85403726708074, Kappa= 0.5608286131339957\n",
      "Checkpoint Saved at 7\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:23<00:00,  1.86it/s, loss=0.393]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 8, Loss= 0.3852541551816683, Accuracy= 85.82624272279445, Kappa= 0.7896977963560072\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.31it/s, loss=1.05] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 8, Loss= 0.7879875062919054, Accuracy= 75.18115942028986, Kappa= 0.5603884537807922\n",
      "Model performance has not improved in this epoch. Patience: 1/3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:22<00:00,  1.86it/s, loss=0.191]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 9, Loss= 0.3589095275216795, Accuracy= 86.92981894952338, Kappa= 0.8096183438700311\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:25<00:00,  2.35it/s, loss=1.07] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 9, Loss= 0.8124582327780177, Accuracy= 74.71532091097309, Kappa= 0.558152608294915\n",
      "Model performance has not improved in this epoch. Patience: 2/3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:29<00:00,  1.81it/s, loss=0.25] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 10, Loss= 0.34280140643105184, Accuracy= 87.40643592860341, Kappa= 0.8224655048280176\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.34it/s, loss=1.09] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 10, Loss= 0.8383681485887433, Accuracy= 74.45652173913044, Kappa= 0.5522528456391598\n",
      "Model performance has not improved in this epoch. Patience: 3/3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 489/489 [04:23<00:00,  1.86it/s, loss=0.446]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Train): 11, Loss= 0.3263430513731054, Accuracy= 88.18373744482119, Kappa= 0.8336561429768927\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 61/61 [00:26<00:00,  2.32it/s, loss=1.13] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch (Val): 11, Loss= 0.8670934624359256, Accuracy= 74.4824016563147, Kappa= 0.5529974630001833\n",
      "Model performance has not improved in this epoch. Patience: 4/3\n",
      "Early Stopping with 0.5608286131339957 at Epoch 7\n"
     ]
    },
    {
     "ename": "",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31mThe Kernel crashed while executing code in the the current cell or a previous cell. Please review the code in the cell(s) to identify a possible cause of the failure. Click <a href='https://aka.ms/vscodeJupyterKernelCrash'>here</a> for more info. View Jupyter <a href='command:jupyter.viewOutput'>log</a> for further details."
     ]
    }
   ],
   "source": [
    "trainedModel = trainModel(model,train_loader,val_loader,optimiser,lr_scheduler1,loss_func,40)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.8.13",
   "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": "3.8.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "83bec4cd18966703d39e3fd8209371b7574445e5ac2dfe27a80cb33e9b531167"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}