liveBook · Manning

Fine-tuning

As explained in Outlier Detection in Python

Fine-tuning is the continued training of a pretrained model, usually on a new task or domain, by adjusting some or all of its parameters.

Fine-tuning continues training a pretrained neural network for a related task, usually by adjusting only some parameters. A model trained to distinguish cats from dogs can be adapted to distinguish different types of dogs. section 16.5.3

The method reuses features or embeddings already learned by a foundation model. Retraining may therefore be limited to the final layer or an added layer, reducing training time, cost, and labeled-data requirements. section 16.5.3

Image classification

Pretrain a network on a broad image task, then fine-tune its representation for a narrower dog-classification task instead of training the second model from scratch. section 16.5.3

Fine-tuning transfers more readily across images than tables because image models learn reusable features such as edges, shapes, and shadows, while tables often differ fundamentally. section 16.5.3

How Outlier Detection in Python covers it

How deep it goes
Explains reused embeddings, selective layer retraining, benefits, and weaker transferability for tabular data.
Examples
Uses cat-versus-dog classification adapted to distinguishing different types of dogs.
What it assumes
Assumes a pretrained neural network and distinguishes its learned embeddings from the target task.
What you can do after
Apply it via a pretrained neural network by retraining a final layer or adding a task-specific layer.
How it connects
Connects transfer learning with foundation models, embeddings, self-supervised learning, and neural-network layers.
Cover of Outlier Detection in Python
liveBookmanning.com
Read Outlier Detection in Python on liveBook Get the book at manning.com

Also explained in

Compare how 37 books explain fine-tuning →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Outlier Detection in Python. The section links point to the parts of the book it draws on.