What is PEFT?

PEFT (Parameter-Efficient Fine-Tuning) is a family of methods that adapts a pretrained model by training selected existing parameters or a comparatively small set of added parameters while most base weights remain frozen. Parameter efficiency does not by itself guarantee a fixed memory reduction or the quality of full fine-tuning.

Quick Facts

Full NameParameter-Efficient Fine-Tuning
CreatedVarious techniques from 2019-2023, unified by Hugging Face
SpecificationOfficial Specification

How It Works

PEFT is a method family, not one algorithm and not a synonym for the Hugging Face PEFT library. Additive methods insert bottleneck modules or learned soft prompts; selective methods update chosen existing parameters; reparameterized methods such as LoRA express an update through smaller trainable factors; hybrid approaches combine mechanisms or quantize the frozen base. These choices affect trainable state, activation memory, optimizer state, context use, inference work, mergeability, and backend support differently. A lower trainable-parameter count usually reduces gradient, optimizer, and checkpoint state, but peak training memory still includes base weights, activations, temporary buffers, runtime state, and safety headroom. Quality must be measured against the unchanged base and a full-tuning candidate when that comparison matters. A PEFT checkpoint is normally a base-dependent artifact rather than a standalone model, so release records need the immutable base, tokenizer, template, method configuration, data revisions, evaluation report, serving composition, and rollback target.

Key Characteristics

  • Covers additive, selective, reparameterized, and hybrid adaptation methods rather than one fixed architecture
  • Keeps most base parameters frozen while training a method-specific subset or newly introduced state
  • Reduces trainable and checkpoint state, but peak memory depends on weights, activations, sequence shape, precision, optimizer, and runtime
  • Produces behavior that must be evaluated by task slice, regression slice, safety policy, and repeated seed
  • Usually creates a base-dependent checkpoint whose compatibility includes model, tokenizer, template, framework, and method configuration
  • Has method-specific serving behavior: some updates can be merged, while modules and soft prompts may require runtime composition

Common Use Cases

  1. Comparing efficient adaptation methods under a fixed workload, device envelope, and evaluation protocol
  2. Maintaining independently versioned domain or task variants on an immutable shared base
  3. Running controlled adaptation experiments without storing a full model copy for every candidate
  4. Adapting language, vision, speech, or diffusion models when the framework supports the selected method
  5. Serving approved model variants with explicit compatibility checks, authorization, observability, and rollback

Example

loading...
Loading code...

Frequently Asked Questions

What is the difference between PEFT and full fine-tuning?

Full fine-tuning updates all or most base-model weights. PEFT trains selected existing parameters or added state while most base weights remain frozen. That usually reduces gradient, optimizer, and checkpoint state, but the actual quality, wall time, and peak-memory difference depends on the model, method, precision, sequence shape, optimizer, runtime, and device topology.

How should a PEFT method be selected?

Start with a workload contract: target behavior, base architecture, allowed training and serving backends, device envelope, latency, storage, and release constraints. Compare viable methods under the same data splits and evaluation protocol. LoRA is broadly supported, bottleneck modules add sequential computation, and soft-prompt methods change attention inputs; none is universally best.

Does training fewer parameters reduce peak memory by the same ratio?

No. Trainable parameters determine only part of training memory. Base weights, quantization metadata, activations, gradients, optimizer state, temporary kernels, communication buffers, runtime allocations, fragmentation, and headroom all contribute. Measure peak device and host memory on the intended sequence length, batch shape, precision, checkpointing policy, and hardware.

Does PEFT always match full fine-tuning quality?

No. Paper results are tied to their models, tasks, data, method configurations, and evaluation protocols. A PEFT candidate can outperform, match, or trail full fine-tuning, and it can improve one slice while regressing another. Compare candidates with repeated seeds, held-out task slices, general-capability regressions, safety checks, and serving-artifact tests.

Is a PEFT checkpoint a standalone model?

Usually not. A PEFT checkpoint commonly stores method-specific parameters and configuration and still requires the compatible base model. Treat the base revision and hash, tokenizer, chat template, framework versions, method configuration, adapter revision, evaluation report, and serving composition as one identity. Merging is method- and configuration-specific and creates a new artifact that needs validation.

Related Terms

LoRA

LoRA (Low-Rank Adaptation) is a parameter-efficient adaptation method that keeps selected pretrained weight matrices frozen and learns additive updates through smaller low-rank matrices. It reduces the trainable update surface, but it does not guarantee a fixed memory saving, hardware fit, training speed, or quality relative to full fine-tuning.

QLoRA

QLoRA (Quantized Low-Rank Adaptation) is a parameter-efficient fine-tuning method that keeps a pretrained base model frozen in 4-bit quantized storage while backpropagating through its dequantized computations into trainable LoRA adapters. It reduces base-weight memory but does not make all training state 4-bit or guarantee that a model will fit a specific device.

Adapter

Adapter is base-dependent trainable state that changes a pretrained model's behavior while most base parameters remain frozen. The term can mean an inserted bottleneck module, a broader PEFT component such as LoRA, or the checkpoint artifact that stores the learned adaptation; these meanings should not be treated as identical.

Fine-tuning

Fine-tuning is the process of continuing optimization from a pretrained model checkpoint on a target dataset and objective, updating some or all model parameters to change task, domain, or behavioral performance.

Related Articles

LoRA Fine-Tuning Tutorial: QLoRA & PEFT Guide (2026)

Learn LoRA fine-tuning step by step with PEFT and QLoRA. Configure rank, alpha, target modules, memory use, adapter merging, and deployment for production LLMs.

2026-02-21

LLM Fine-Tuning【2026】: SFT, LoRA, QLoRA, and Evaluation

A rigorous guide to adapting language models with supervised fine-tuning and parameter-efficient methods. Learn when training beats prompting or RAG, how to build a licensed and leakage-resistant dataset, estimate memory instead of repeating hardware folklore, run version-pinned experiments, and evaluate capability, safety, regression, and uncertainty.

2026-02-21

DPO vs RLHF: The Evolution of LLM Alignment Techniques

A deep technical comparison of DPO and RLHF for LLM alignment. Covers reward model training, PPO instabilities, the Bradley-Terry framework behind DPO, compute costs, and newer variants like KTO, IPO, ORPO, and SimPO.

2026-04-23