import torch
import torchvision
from torchvision.models import resnet50
from torchvision.transforms import transforms
from torch.utils.data import DataLoader, Dataset

# Define the dataset class
class RobloxDataset(Dataset):
    def __init__(self, root_dir, transform=None):
        self.root_dir = root_dir
        self.transform = transform
        
    def __len__(self):
        return 200  # Number of images in the dataset, replace with your own to train with your own images.
    
    def __getitem__(self, idx):
        img_path = f'{self.root_dir}/human_{str(idx+1).zfill(2)}.png'
        image = Image.open(img_path).convert('RGB')
        
        if self.transform:
            image = self.transform(image)
        
        return image

data_transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

dataset = RobloxDataset('/content/dataset', transform=data_transform)

data_loader = DataLoader(dataset, batch_size=1, shuffle=True)

model = resnet50(pretrained=True)
model.fc = torch.nn.Linear(in_features=2048, out_features=1)  # Adjust the number of output classes if needed

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
model.to(device)

criterion = torch.nn.BCEWithLogitsLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)

# Train the model
num_epochs = 100  # Adjust the number of training epochs (more epochs = more training time but with more accuracy and less loss).
""" Training the model with more epochs
Pros:
- more accuracy
- less loss (it means the model is improving)

Con:
- more training time
"""

for epoch in range(num_epochs):
    for images in data_loader:
        images = images.to(device)
        labels = torch.ones((images.size(0), 1)).to(device)  # Assuming all images belong to the same class
 
        outputs = model(images)
        loss = criterion(outputs, labels)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        
    print(f'Epoch [{epoch+1}/{num_epochs}], Loss: {loss.item()}')

# Save the trained model
torch.save(model.state_dict(), '/content/zero_shot_classification_model.pth')