Refactor training loop from script to class

README.md

@@ -1 +1,33 @@
-# evals
+# evals
+
+This project contains a PyTorch implementation for training and evaluating a simple neural network on the MNIST dataset.
+
+## ModelTrainer class
+
+The `ModelTrainer` class is used to train and evaluate the model. Here is how you can use it:
+
+```python
+from main import ModelTrainer
+
+# Create an instance of ModelTrainer
+trainer = ModelTrainer()
+
+# Run the training and evaluation process
+trainer.run()
+```
+
+The `ModelTrainer` class has the following methods:
+
+- `load_data()`: Loads the MNIST dataset and applies the necessary transformations.
+- `define_model()`: Defines the model architecture.
+- `train_model()`: Trains the model on the MNIST dataset.
+- `evaluate()`: Evaluates the model performance on the test dataset.
+- `run()`: Runs the entire process (data loading, model definition, training, and evaluation).
+
+## Saving and Loading the Model
+
+After training, the model's parameters are saved to a file named "mnist_model.pth". You can load the model parameters using the `torch.load()` function:
+
+```python
+model.load_state_dict(torch.load("mnist_model.pth"))
+```

src/api.py

@@ -2,12 +2,11 @@
 from PIL import Image
 import torch
 from torchvision import transforms
-from main import Net  # Importing Net class from main.py
+from main import ModelTrainer  # Importing ModelTrainer class from main.py
 
-# Load the model
-model = Net()
-model.load_state_dict(torch.load("mnist_model.pth"))
-model.eval()
+# Create an instance of ModelTrainer and load the model
+trainer = ModelTrainer()
+model = trainer.load_model("mnist_model.pth")
 
 # Transform used for preprocessing the image
 transform = transforms.Compose([

src/main.py

@@ -6,43 +6,80 @@
 from torch.utils.data import DataLoader
 import numpy as np
 
-# Step 1: Load MNIST Data and Preprocess
-transform = transforms.Compose([
-    transforms.ToTensor(),
-    transforms.Normalize((0.5,), (0.5,))
-])
-
-trainset = datasets.MNIST('.', download=True, train=True, transform=transform)
-trainloader = DataLoader(trainset, batch_size=64, shuffle=True)
-
-# Step 2: Define the PyTorch Model
-class Net(nn.Module):
-    def __init__(self):
-        super().__init__()
-        self.fc1 = nn.Linear(28 * 28, 128)
-        self.fc2 = nn.Linear(128, 64)
-        self.fc3 = nn.Linear(64, 10)
-
-    def forward(self, x):
-        x = x.view(-1, 28 * 28)
-        x = nn.functional.relu(self.fc1(x))
-        x = nn.functional.relu(self.fc2(x))
-        x = self.fc3(x)
-        return nn.functional.log_softmax(x, dim=1)
-
-# Step 3: Train the Model
-model = Net()
-optimizer = optim.SGD(model.parameters(), lr=0.01)
-criterion = nn.NLLLoss()
-
-# Training loop
-epochs = 3
-for epoch in range(epochs):
-    for images, labels in trainloader:
-        optimizer.zero_grad()
-        output = model(images)
-        loss = criterion(output, labels)
-        loss.backward()
-        optimizer.step()
-
-torch.save(model.state_dict(), "mnist_model.pth")
+class ModelTrainer:
+    def __init__(self, epochs=3, lr=0.01, batch_size=64):
+        self.epochs = epochs
+        self.lr = lr
+        self.batch_size = batch_size
+        self.model = None
+        self.optimizer = None
+        self.criterion = nn.NLLLoss()
+        self.trainloader = None
+
+    def load_data(self):
+        transform = transforms.Compose([
+            transforms.ToTensor(),
+            transforms.Normalize((0.5,), (0.5,))
+        ])
+
+        trainset = datasets.MNIST('.', download=True, train=True, transform=transform)
+        self.trainloader = DataLoader(trainset, batch_size=self.batch_size, shuffle=True)
+
+    def define_model(self):
+        class Net(nn.Module):
+            def __init__(self):
+                super().__init__()
+                self.fc1 = nn.Linear(28 * 28, 128)
+                self.fc2 = nn.Linear(128, 64)
+                self.fc3 = nn.Linear(64, 10)
+
+            def forward(self, x):
+                x = x.view(-1, 28 * 28)
+                x = nn.functional.relu(self.fc1(x))
+                x = nn.functional.relu(self.fc2(x))
+                x = self.fc3(x)
+                return nn.functional.log_softmax(x, dim=1)
+
+        self.model = Net()
+        self.optimizer = optim.SGD(self.model.parameters(), lr=self.lr)
+
+    def train_model(self):
+        for epoch in range(self.epochs):
+            for images, labels in self.trainloader:
+                self.optimizer.zero_grad()
+                output = self.model(images)
+                loss = self.criterion(output, labels)
+                loss.backward()
+                self.optimizer.step()
+
+    def evaluate(self):
+        transform = transforms.Compose([
+            transforms.ToTensor(),
+            transforms.Normalize((0.5,), (0.5,))
+        ])
+
+        testset = datasets.MNIST('.', download=True, train=False, transform=transform)
+        testloader = DataLoader(testset, batch_size=self.batch_size, shuffle=True)
+
+        correct = 0
+        total = 0
+
+        with torch.no_grad():
+            for images, labels in testloader:
+                output = self.model(images)
+                _, predicted = torch.max(output.data, 1)
+                total += labels.size(0)
+                correct += (predicted == labels).sum().item()
+
+        accuracy = 100 * correct / total
+        print('Accuracy of the model on the test images: %d%%' % accuracy)
+
+    def run(self):
+        self.load_data()
+        self.define_model()
+        self.train_model()
+        self.evaluate()
+        torch.save(self.model.state_dict(), "mnist_model.pth")
+
+trainer = ModelTrainer()
+trainer.run()