finish chapter 15

.gitignore

@@ -5,5 +5,5 @@ lib64
 share/
 pyvenv.cfg
 .python-version
-data/MNIST/
+data/*
-data/CIFAR10/
+!data/wine/

14_cnn.py

@@ -27,8 +27,8 @@ device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 print(f'Device is {device}')
 
 # hyper parameters
-num_epochs = 4
+num_epochs = 10
-batch_size = 4
+batch_size = 10
 learning_rate = 0.001
 
 # dataset has PILImage images of range [0,1]
@@ -51,24 +51,44 @@ test_loader = torch.utils.data.DataLoader(test_dataset, batch_size=batch_size)
 classes = ('plane', 'car', 'bird', 'cat', 'deer', 
            'dog', 'frog', 'horse', 'ship', 'truck') # can also get them from the data, probably
 
-#implement conv net
-class ConvNet(nn.Module):
-    def __init__(self) -> None:
-        super().__init__()
-        self.layers = nn.Sequential([
-            nn.Conv2d(3, 6, 5), # 3 color channels, 6 output channels, kernel size 5
-            nn.MaxPool2d(2, 2), # kernel size 2, stride 2 (whats stride?)
-            
-            # 3:23:19
 
-        ])
+# #implement conv net
-        pass
+# class ConvNet(nn.Module):
+#     def __init__(self):
+#         super(ConvNet, self).__init__()
+#         self.layers = nn.Sequential(                                                                                                   # -> 32x32
+#             nn.Conv2d(3, 6, 5), # 3 color channels, 6 output channels, kernel size 5 -> image size shrinks by 2 pixels in each direction -> 28x28
+#             nn.ReLU(),
+#             nn.MaxPool2d(2, 2), # kernel size 2, stride 2 -> shift by 2 pixels after each max-pooling -> image size shrinks to half size -> 14x14
+#             nn.Conv2d(6, 16, 5), # input size is output size of previous conv layer, -> image size shrinks by 2 pixels in each direction -> 10x10
+#             nn.ReLU(),
+#             nn.MaxPool2d(2, 2), # kernel size 2, stride 2 -> shift by 2 pixels after each max-pooling -> image size shrinks to half size ->  5x5
+#             nn.Flatten(),
+#             nn.Linear(16*5*5, 120), # 16 channels * 5px * 5px
+#             nn.ReLU(),
+#             nn.Linear(120, 84),
+#             nn.ReLU(),
+#             nn.Linear(84, 10) # output size 10 for 10 classes
+#         )
     
-    def forward(self, x):
+#     def forward(self, x):
-        # return self.layers(x)
+#         return self.layers(x)
-        pass
 
-model = ConvNet().to(device)
+# model = ConvNet().to(device)
+model = nn.Sequential(                                                                                                   # -> 32x32
+            nn.Conv2d(3, 6, 5), # 3 color channels, 6 output channels, kernel size 5 -> image size shrinks by 2 pixels in each direction -> 28x28
+            nn.ReLU(),
+            nn.MaxPool2d(2, 2), # kernel size 2, stride 2 -> shift by 2 pixels after each max-pooling -> image size shrinks to half size -> 14x14
+            nn.Conv2d(6, 16, 5), # input size is output size of previous conv layer, -> image size shrinks by 2 pixels in each direction -> 10x10
+            nn.ReLU(),
+            nn.MaxPool2d(2, 2), # kernel size 2, stride 2 -> shift by 2 pixels after each max-pooling -> image size shrinks to half size ->  5x5
+            nn.Flatten(),
+            nn.Linear(16*5*5, 120), # 16 channels * 5px * 5px
+            nn.ReLU(),
+            nn.Linear(120, 84),
+            nn.ReLU(),
+            nn.Linear(84, 10) # output size 10 for 10 classes
+        ).to(device)
 
 criterion = nn.CrossEntropyLoss()
 optimizer = torch.optim.SGD(model.parameters(), lr = learning_rate)

15_transfer.py

@@ -0,0 +1,153 @@
+# new:
+
+import torch
+import torch.nn as nn
+import torch.optim as optim
+from torch.optim import lr_scheduler
+import numpy as np
+import torchvision
+from torchvision import datasets, models, transforms
+import matplotlib.pyplot as plt
+import time
+import os
+import copy
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+mean = np.array([0.485, 0.456, 0.406])
+std = np.array([0.229, 0.224, 0.225])
+
+data_transforms = {
+    'train': transforms.Compose([
+        transforms.RandomResizedCrop(224),
+        transforms.RandomHorizontalFlip(),
+        transforms.ToTensor(),
+        transforms.Normalize(mean, std)
+    ]),
+    'val': transforms.Compose([
+        transforms.Resize(256),
+        transforms.CenterCrop(224),
+        transforms.ToTensor(),
+        transforms.Normalize(mean, std)
+    ])
+}
+
+#import data
+data_dir = 'data/hymenoptera_data'
+sets = ['train', 'val']
+image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), data_transforms[x]) for x in sets}
+dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4, shuffle=True, num_workers=0) for x in sets}
+dataset_sizes = {x: len(image_datasets[x]) for x in sets}
+class_names = image_datasets['train'].classes
+print(f'Classes: {class_names}')
+
+
+# # show sample images
+# def imshow(inp, title):
+#     """Imshow for Tensor."""
+#     inp = inp.numpy().transpose((1, 2, 0))
+#     inp = std * inp + mean
+#     inp = np.clip(inp, 0, 1)
+#     plt.imshow(inp)
+#     plt.title(title)
+#     plt.show()
+# inputs, classes = next(iter(dataloaders['train']))
+# out = torchvision.utils.make_grid(inputs)
+# imshow(out, title=[class_names[x] for x in classes])
+
+
+def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
+    since = time.time()
+    best_model_wts = copy.deepcopy(model.state_dict())
+    best_acc = 0.0
+
+    for epoch in range(num_epochs):
+        print(F'Epoch {epoch+1}/{num_epochs}')
+        print('-'*11)
+
+        for phase in sets:
+            if phase == 'train':
+                model.train()
+            else:
+                model.eval()
+            
+            running_loss = 0.0
+            running_corrects = 0
+
+            for inputs, labels in dataloaders[phase]:
+                inputs = inputs.to(device)
+                labels = labels.to(device)
+
+                with torch.set_grad_enabled(phase == 'train'):
+                    outputs = model(inputs)
+                    _, preds = torch.max(outputs, 1)
+                    loss = criterion(outputs, labels)
+
+                    if phase == 'train':
+                        optimizer.zero_grad()
+                        loss.backward()
+                        optimizer.step()
+                
+                running_loss += loss.item() * inputs.size(0)
+                running_corrects += torch.sum(preds == labels.data)
+            
+            if phase == 'train':
+                scheduler.step()
+            
+            epoch_loss = running_loss/dataset_sizes[phase]
+            epoch_acc = running_corrects.double()/dataset_sizes[phase]
+
+            print(f'{phase} Loss: {epoch_loss:.4f}, Acc: {epoch_acc:.4f}')
+
+            if phase == 'val' and epoch_acc > best_acc:
+                best_acc = epoch_acc
+                best_model_wts = copy.deepcopy(model.state_dict())
+            
+        print()
+    
+    time_elapsed = time.time() - since
+    print(f'Training complete in {time_elapsed//(60*60):02.0f}:{(time_elapsed%(60*60)//60):02.0f}:{time_elapsed%60:02.0f}')
+    print(f'Best val Acc: {best_acc:.4f}')
+
+    model.load_state_dict(best_model_wts)
+    return model
+
+## transfer learning - finetuning ##
+model = torchvision.models.resnet18(weights=torchvision.models.ResNet18_Weights.DEFAULT) # import pretrained model
+
+#change last layer
+num_ftrs = model.fc.in_features
+model.fc = nn.Linear(num_ftrs, 2)
+model.to(device)
+
+criterion = nn.CrossEntropyLoss()
+optimizer = torch.optim.SGD(model.parameters(), lr=0.001)
+
+# scheduler
+# updates the learning rate
+step_lr_scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=7, gamma=0.1) # every seven epochs: lr = lr*gamma
+
+model = train_model(model, criterion, optimizer, step_lr_scheduler, num_epochs=50)
+
+print('-'*50)
+print()
+
+## transfer learning - feature extractor ##
+model = torchvision.models.resnet18(weights=torchvision.models.ResNet18_Weights.DEFAULT) # import pretrained model # import pretrained model
+for param in model.parameters():
+    param.requires_grad = False
+    # freezes all layers in beginning
+
+#change last layer
+num_ftrs = model.fc.in_features
+model.fc = nn.Linear(num_ftrs, 2)
+model.to(device)
+
+criterion = nn.CrossEntropyLoss()
+optimizer = torch.optim.SGD(model.parameters(), lr=0.001)
+
+# scheduler
+# updates the learning rate
+step_lr_scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=7, gamma=0.1) # every seven epochs: lr = lr*gamma
+
+model = train_model(model, criterion, optimizer, step_lr_scheduler, num_epochs=50)