How deeply does The Art of AI Product Development cover parameter-efficient fine-tuning?

Defines PEFT by its limited parameter updates and compares its costs with full fine-tuning, without specifying an implementation variant.

Parameter-efficient fine-tuning

As explained in The Art of AI Product Development

Parameter-efficient fine-tuning is a fine-tuning method that adapts a model by updating a small subset or additional parameterization rather than all model parameters.

Parameter-efficient fine-tuning (PEFT) adapts a language model by updating only a small subset of its parameters rather than all of them. Compared with full fine-tuning, this makes customization faster and less resource-intensive—especially when a product needs repeated updates or several specialized models. section 8.2.1 section 8.2.2

Scaling client-specific classification

A content application fine-tunes a small language model to classify documents by topic so generated content draws from relevant sources. Its pilot reaches 92% accuracy, but training a separate full model for every client would be difficult to scale. PEFT makes it more practical to maintain models reflecting different client domains and taxonomies. Users can flag incorrect source choices, which become corrections for later training. section 8.2.2

When deciding whether PEFT fits, check:

Model access and compatibility. Confirm that the candidate model supports PEFT and permits parameter changes; inference-only access is not enough for this approach. section 8.2.1
The quality–cost trade-off. Test promising model sizes with small, representative datasets instead of committing early to the largest candidate. section 8.2.1
The improvement loop. Track errors, add corrective examples, and evaluate whether customization improves the application rather than assuming every new technique will help. section 5.4.5

PEFT reduces the cost of changing model parameters; it does not replace good data, application-specific evaluation, or user feedback. Treat it as one option alongside prompting and retrieval improvements, chosen for measurable benefit rather than novelty. section 5.4.5 section 8.3

How The Art of AI Product Development covers it

Cover of The Art of AI Product Development

liveBook manning.com

Read The Art of AI Product Development on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

LLMs in ProductionParameter-efficient fine-tuning is a fine-tuning approach that adapts a model by training only a small subset of parameters or added parameters.
liveBook manning.com

Domain-Specific Small Language ModelsA family of fine-tuning methods that adapt a pretrained model using only a small subset of parameters or added modules to reduce compute and cost.
liveBook manning.com

Financial AI in PracticeParameter-efficient fine-tuning adapts a pretrained model by training a small subset of its parameters or adding lightweight trainable components.
liveBook manning.com

Building Reliable AI SystemsFine-tuning methods that adapt a small subset or representation of a model’s parameters to reduce training resource requirements.

Compare how 5 books explain parameter-efficient fine-tuning →

Related concepts

low-rank adaptationincludes · Low-rank adaptation is a parameter-efficient fine-tuning method that adapts a model by learning low-rank updates to its weights, typically while keeping most original weights fixed.
fine-tuningkind of · Fine-tuning is adapting a pretrained machine-learning model by training it further on targeted data or objectives for a task, domain, or behavior.
QLoRAincludes · QLoRA is a parameter-efficient fine-tuning method that quantizes a pretrained model’s weights while training low-rank adapters to reduce memory requirements.
catastrophic forgettingaddresses · Catastrophic forgetting is the substantial loss of previously learned information or capabilities when a model learns new data, tasks, or objectives.

This explainer was generated by Manning's knowledge graph from the text of The Art of AI Product Development. The section links point to the parts of the book it draws on.