2nd/10_Wiki/Topics/AI_and_ML/PEFT (Parameter-Efficient Fine-Tuning).md

---
id: wiki-2026-0508-peft-parameter-efficient-fine-tu
title: PEFT (Parameter-Efficient Fine-Tuning)
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [PEFT, Parameter-Efficient Fine-Tuning, LoRA fine-tuning]
duplicate_of: none
source_trust_level: A
confidence_score: 0.95
verification_status: applied
tags: [peft, lora, qlora, fine-tuning, llm]
raw_sources: []
last_reinforced: 2026-05-10
github_commit: pending
tech_stack:
  language: python
  framework: peft, transformers, bitsandbytes
---

# PEFT (Parameter-Efficient Fine-Tuning)

## 매 한 줄
> **"매 frozen base + tiny trainable delta"**. 매 full fine-tuning 의 ~0.1-1% parameter 만 학습. 2026 standard: LoRA / QLoRA — 매 70B model 도 single 24GB GPU 에서 fine-tune 가능. HuggingFace `peft` library 의 사실상 표준.

## 매 핵심

### 매 동기
- Full FT: 70B model = 280GB (fp32) gradient + optimizer state → multi-A100 cluster 필요.
- Storage: 매 task 마다 full checkpoint 저장 시 비용 폭발.
- Catastrophic forgetting: 매 full FT 가 base capability 손상.
- PEFT: 매 base frozen, delta 만 학습 → 1 base + N tiny adapters.

### 매 family
- **LoRA** (Hu et al. 2021): low-rank decomposition `ΔW = BA`, rank r=4-64.
- **QLoRA** (Dettmers et al. 2023): 4-bit NF4 quantized base + LoRA adapters.
- **Prefix Tuning** (Li & Liang 2021): learnable prefix tokens prepended to keys/values.
- **Prompt Tuning** (Lester et al. 2021): learnable soft prompts at input.
- **IA³** (Liu et al. 2022): scale activations via learned vectors (multiply, not add).
- **Adapters** (Houlsby et al. 2019): small bottleneck MLPs inserted between layers.
- **DoRA** (2024): magnitude + direction decomposition, LoRA 보다 우수.

### 매 LoRA 수학
- `W' = W + αBA/r` where `B ∈ R^{d×r}`, `A ∈ R^{r×k}`, r ≪ min(d,k).
- Trainable: `2dr` params instead of `dk`. 매 d=k=4096, r=8 → 65k vs 16M (250× 감소).
- Inference: 매 merge `W ← W + αBA/r` → zero overhead.

### 매 응용
1. Domain adaptation (legal, medical LLM).
2. Instruction tuning (Alpaca-style).
3. Style transfer (FLUX LoRA for art style).
4. Multi-tenant serving (1 base + N customer LoRAs).

## 💻 패턴

### LoRA with peft library
```python
from peft import LoraConfig, get_peft_model, TaskType
from transformers import AutoModelForCausalLM

base = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B")
config = LoraConfig(
    task_type=TaskType.CAUSAL_LM,
    r=16, lora_alpha=32, lora_dropout=0.05,
    target_modules=["q_proj", "k_proj", "v_proj", "o_proj"],
)
model = get_peft_model(base, config)
model.print_trainable_parameters()  # ~0.5% trainable
```

### QLoRA (4-bit base + LoRA)
```python
from transformers import BitsAndBytesConfig
import torch

bnb = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.bfloat16,
    bnb_4bit_use_double_quant=True,
)
base = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-3.1-70B", quantization_config=bnb, device_map="auto",
)
model = get_peft_model(base, lora_config)  # 70B on single 48GB GPU
```

### Save / load adapter only
```python
model.save_pretrained("./my-lora")  # ~50MB, not 140GB

from peft import PeftModel
loaded = PeftModel.from_pretrained(base, "./my-lora")
```

### Merge for inference
```python
merged = model.merge_and_unload()  # W ← W + αBA/r
merged.save_pretrained("./merged-model")  # standard HF model, no peft dep
```

### Multi-LoRA serving (vLLM)
```python
# vLLM 0.6+ supports dynamic LoRA loading
from vllm import LLM, SamplingParams
from vllm.lora.request import LoRARequest

llm = LLM(model="meta-llama/Llama-3.1-8B", enable_lora=True, max_loras=8)
out = llm.generate(prompts, sampling_params,
    lora_request=LoRARequest("customer-42", 1, "./customer-42-lora"))
```

### DoRA (2024)
```python
config = LoraConfig(r=16, lora_alpha=32, use_dora=True,  # peft >= 0.10
    target_modules=["q_proj", "v_proj"])
```

### Prompt tuning
```python
from peft import PromptTuningConfig, PromptTuningInit
config = PromptTuningConfig(
    task_type=TaskType.CAUSAL_LM,
    prompt_tuning_init=PromptTuningInit.TEXT,
    num_virtual_tokens=20,
    prompt_tuning_init_text="Classify sentiment:",
    tokenizer_name_or_path="meta-llama/Llama-3.1-8B",
)
```

## 매 결정 기준
| 상황 | Approach |
|---|---|
| 1 GPU, large base (70B) | QLoRA |
| Multi-task, single base | LoRA + multi-adapter serving |
| Tiny VRAM, frozen base OK | Prompt tuning |
| Best quality, less compute saving | DoRA |
| Diffusion model style | LoRA (rank 4-32) |
| Production accuracy critical | Full FT (if 가능) |

**기본값**: QLoRA (4-bit NF4 + r=16 LoRA on q/k/v/o projections).

## 🔗 Graph
- 부모: [[Fine-Tuning]]
- 변형: [[LoRA]] · [[QLoRA]] · [[DoRA]]
- 응용: [[Fine-tuning|Instruction-Tuning]] · [[Domain-Adaptation]]
- Adjacent: [[LLM_Optimization_and_Deployment_Strategies|Quantization]] · [[RLHF]]

## 🤖 LLM 활용
**언제**: 매 single GPU 에서 large model fine-tune, multi-tenant LoRA serving, rapid task iteration.
**언제 X**: 매 base model 의 fundamental capability 변경 필요 (continued pretraining → full FT or full pretraining).

## ❌ 안티패턴
- **Rank too low**: r=1-2 → underfitting. 매 r=8-32 starting point.
- **Wrong target modules**: only `q_proj`/`v_proj` skip → degraded. 매 all attention + MLP modules 가 best.
- **Forgetting alpha**: 매 alpha=2r convention 무시 → unstable training.
- **Saving full model**: `model.save_pretrained()` on PeftModel 만 saves adapter. Don't merge unnecessarily.
- **QLoRA + bf16 base**: 매 NF4 quantization 의 redundant. 매 fp16 or bf16 base 둘 중 하나.

## 🧪 검증 / 중복
- Verified (HuggingFace `peft` docs, Hu et al. 2021 LoRA, Dettmers et al. 2023 QLoRA).
- 신뢰도 A.

## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — PEFT family, LoRA/QLoRA patterns, decision matrix |