How can I fine-tune a pretrained model in PyTorch?

How can I fine-tune a pretrained model in PyTorch?

To fine-tune a pretrained model in PyTorch, you can start by loading a model from torchvision, like ResNet. Then, modify the final layer to match your number of classes.

What steps do I need to take to prepare my dataset?

You need to preprocess your dataset, which includes resizing images, normalizing them, and creating data loaders for training and validation.

How do I set the learning rate for fine-tuning?

Typically, a lower learning rate is used for fine-tuning, often around 1e-4 or 1e-5. You can adjust it based on your results.

Can you show me a code example for fine-tuning?

Sure! Here's a simple example:

import torch
import torchvision.models as models

model = models.resnet18(pretrained=True)
model.fc = torch.nn.Linear(model.fc.in_features, num_classes)

# Set the model to training mode
model.train()