Transfer learning

As explained in LLMs in Production

Transfer learning is a machine-learning technique that adapts knowledge learned from one task or dataset to another task or domain.

Suppose a general-purpose language model needs to answer questions in a particular domain. Training a new model from scratch would repeat the expensive work of learning broad language patterns. Transfer learning reuses knowledge learned for a source problem and adapts it to a related target task or domain. For LLMs, that usually means starting with a pretrained model and finetuning it on a smaller, task-specific dataset. section 5.2.2

The practical sequence is:

• Start with a foundation model pretrained on a large, diverse corpus. Its existing weights capture general language patterns; creating those weights requires substantial computation. section 5.2.2
• Update the model using data representative of the target task or domain. This generally takes far less data and computation than pretraining, and can improve performance on tasks such as sentiment analysis, text classification, or question answering. section 2.3.3 section 5.2.2
• Consider prompting when examples in a query are enough. Large models can sometimes perform specialized tasks without finetuning, so updating weights is not the only route to adaptation. section 2.4

Choose the adaptation path

A pretrained BERT model can be retrained on data representative of a particular task instead of being trained from scratch. For an LLM, the same general-to-specific pattern can produce an expert-task model from a foundation model and a small, curated dataset. If a few examples in the prompt already produce suitable results, prompting may avoid that finetuning step. section 2.3.3 section 5.2 section 2.4

Figure 5.3 Finetuning differs from training from scratch in that you don’t have to define model behavior, you can use the exact same training loop, and you have a fraction of the data requirement.

The comparison makes the central resource trade-off visible: finetuning reuses an existing model and needs a fraction of the data required for training from scratch. section 5.2.2

How LLMs in Production covers it

How deep it goes

Defines source and target domains, explains pretraining followed by finetuning, and contrasts training costs without detailing weight-update internals.

Examples

Uses BERT task adaptation, sentiment analysis, text classification, question answering, and a finetuning-versus-scratch comparison.

What it assumes

Defines source and target domains and places adaptation after pretraining, without requiring the reader to create a foundation model.

What you can do after

Apply it via OpenAI fine-tuning jobs or a pretrained model, using target-specific data rather than training from scratch.

How it connects

Connects foundation models and finetuning to pretraining, curated datasets, and few-shot prompting.

liveBookmanning.com

Read LLMs in Production on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Inside Deep Learning: Math, Algorithms, ModelsA training approach that adapts a pretrained model to a new task, usually by reusing learned weights and matching the original preprocessing expectations.
• 

  liveBookmanning.com
  Natural Language Processing in Action, Second EditionTransfer learning is reusing knowledge learned from one task or dataset to improve learning on another task or dataset.
• 

  liveBookmanning.com
  Generative AI in ActionTransfer learning adapts knowledge learned from one task or dataset to improve performance on another task.
• 

  liveBookmanning.com
  Knowledge Graphs and LLMs in ActionTransfer learning is a machine-learning technique that reuses patterns learned from one task or domain to improve performance on another task or domain.
• 

  liveBookmanning.com
  Time Series Forecasting Using Foundation ModelsTransfer learning is a technique that applies knowledge learned by a model on one dataset or task to a different dataset or task.
• 

  liveBookmanning.com
  Grokking Machine LearningTransfer learning is a technique that adapts a pretrained model to a new task or dataset by continuing its training, often by modifying only some parameters.
• 

  liveBookmanning.com
  Machine Learning for Tabular Data: XGBoost, Deep Learning, and AIA technique that adapts knowledge learned from one task or dataset to a related task.
• 

  liveBookmanning.com
  Sutskever's ListTransfer learning is a technique that adapts knowledge learned from one task or dataset to another task or dataset.
• 

  liveBookmanning.com
  Domain-Specific Small Language ModelsThe process of taking a model trained on one task or domain and applying it to a related task or domain, often with further tuning.
• 

  liveBookmanning.com
  Outlier Detection in PythonTransfer learning is a technique that adapts knowledge learned by a model on one task or dataset to another task or dataset.

Compare how 11 books explain transfer learning →

Related concepts

• fine-tuningused by · Fine-tuning is adapting a pretrained machine-learning model by training it further on targeted data or objectives for a task, domain, or behavior.
• pretrained modeluses · A pretrained model is a machine-learning model whose parameters were learned through prior training before it is used or adapted for a subsequent task.
• embeddinguses · A numerical vector representation of an item or input that encodes features or relationships useful for computational tasks.
• generalizationuses · Generalization is the extension of a pattern, model, solution, or conclusion beyond the specific examples or cases from which it was derived.
• unlabeled datauses · Unlabeled data consists of examples that lack target labels or task-specific annotations.
• causal invarianceused by · Causal invariance is the stability of a causal relationship across changes in environments or data distributions.

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