SKILLoptionalHermes Optional Skills

Peft

Name: Peft
Author: Orchestra Research

Fine-tune large LLMs with LoRA on limited GPU memory

Fine-TuningPEFTLoRAQLoRAParameter-EfficientAdaptersLow-RankMemory Optimization

Last registry verification2026-08-18v1.0.0Orchestra Research

Open original source ↗اقرأ بالعربية

Plain meaning

What does it add to Hermes?

Fine-tune large LLMs with LoRA on limited GPU memory

Peft is a skill related to memory and knowledge. It helps Hermes retrieve past information or stored knowledge instead of starting from zero in every session.

This plain-language explanation is based on the publisher description. The original text remains visible for verification.

Use it when

Use it when your goal in memory and knowledge is clear and you can limit it to the data and actions it actually needs.

Skip it when

Do not add it merely to experiment when Hermes already has a simpler path, or when you cannot review its source and permissions.

Who is it for?

Best for users who want a repeatable way of working inside Hermes.

Safe first test

Add a non-sensitive note, start a new session, and verify retrieval preserves the meaning without inventing details.

Original publisher description

Fine-tune large LLMs with LoRA on limited GPU memory

✓

Data source

This entry was indexed from Hermes Optional Skills. Our explanation interprets the type and domain without inventing a capability not present upstream.

Security review

The source is official or editorially reviewed, but you still need to review permissions and version compatibility.

Safe setup path

Inspect, install, then test.

01
Open the source
Match the publisher, license, and description to your need. Check the real update history.
02
Review permissions and secrets
Never paste a secret value into this site. Use environment-variable names and grant the smallest scope.
03
Copy setup only after review
The controls below copy text. They do not execute commands on your device.
04
Test with a non-sensitive task
Inspect the visible tools, then exclude write or delete tools you do not need.

Install command

Review the command, then copy it.

hermes skills install peft

Hermes Belarabi does not execute this command. Installation happens on your device and remains subject to Hermes scanning and your review.

The full skill definition

Exactly what Hermes loads when this skill runs.

Reproduced from the official documentation. Read it before enabling the skill: this text becomes the agent's instructions.

Fine-tune large LLMs with LoRA on limited GPU memory.

Skill metadata

A lookup table. Do not read it all; find the row that applies to you.

Source	Optional — install with `hermes skills install official/mlops/peft`
Path	`optional-skills/mlops/peft`
Version	`1.0.0`
Author	Orchestra Research
License	MIT
Dependencies	`peft>=0.13.0`, `transformers>=4.45.0`, `torch>=2.0.0`, `bitsandbytes>=0.43.0`
Platforms	linux, macos, windows
Tags	`Fine-Tuning`, `PEFT`, `LoRA`, `QLoRA`, `Parameter-Efficient`, `Adapters`, `Low-Rank`, `Memory Optimization`, `Multi-Adapter`

Reference: full SKILL.md

Explains the idea itself. Read it slowly; the later sections build on it.

Fine-tune LLMs by training <1% of parameters using LoRA, QLoRA, and 25+ adapter methods.

When to use PEFT

Explains the idea itself. Read it slowly; the later sections build on it.

Use PEFT/LoRA when:

Fine-tuning 7B-70B models on consumer GPUs (RTX 4090, A100)
Need to train <1% parameters (6MB adapters vs 14GB full model)
Want fast iteration with multiple task-specific adapters
Deploying multiple fine-tuned variants from one base model

Use QLoRA (PEFT + quantization) when:

Fine-tuning 70B models on single 24GB GPU
Memory is the primary constraint
Can accept ~5% quality trade-off vs full fine-tuning

Use full fine-tuning instead when:

Training small models (<1B parameters)
Need maximum quality and have compute budget
Significant domain shift requires updating all weights

Quick start

Ordered, practical steps. Run one and confirm it worked before moving on.

Installation

Shell8 lines

# Basic installation
pip install peft

# With quantization support (recommended)
pip install peft bitsandbytes

# Full stack
pip install peft transformers accelerate bitsandbytes datasets

LoRA fine-tuning (standard)

Python59 lines

from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, Trainer
from peft import get_peft_model, LoraConfig, TaskType
from datasets import load_dataset

# Load base model
model_name = "meta-llama/Llama-3.1-8B"
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
tokenizer.pad_token = tokenizer.eos_token

# LoRA configuration
lora_config = LoraConfig(
    task_type=TaskType.CAUSAL_LM,
    r=16,                          # Rank (8-64, higher = more capacity)
    lora_alpha=32,                 # Scaling factor (typically 2*r)
    lora_dropout=0.05,             # Dropout for regularization
    target_modules=["q_proj", "v_proj", "k_proj", "o_proj"],  # Attention layers
    bias="none"                    # Don't train biases
)

# Apply LoRA
model = get_peft_model(model, lora_config)
model.print_trainable_parameters()
# Output: trainable params: 13,631,488 || all params: 8,043,307,008 || trainable%: 0.17%

# Prepare dataset
dataset = load_dataset("databricks/databricks-dolly-15k", split="train")

def tokenize(example):
    text = f"### Instruction:\n{example['instruction']}\n\n### Response:\n{example['response']}"
    return tokenizer(text, truncation=True, max_length=512, padding="max_length")

tokenized = dataset.map(tokenize, remove_columns=dataset.column_names)

# Training
training_args = TrainingArguments(
    output_dir="./lora-llama",
    num_train_epochs=3,
    per_device_train_batch_size=4,
    gradient_accumulation_steps=4,
    learning_rate=2e-4,
    fp16=True,
    logging_steps=10,
    save_strategy="epoch"
)

trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized,
    data_collator=lambda data: {"input_ids": torch.stack([f["input_ids"] for f in data]),
                                 "attention_mask": torch.stack([f["attention_mask"] for f in data]),
                                 "labels": torch.stack([f["input_ids"] for f in data])}
)

trainer.train()

# Save adapter only (6MB vs 16GB)
model.save_pretrained("./lora-llama-adapter")

QLoRA fine-tuning (memory-efficient)

Python33 lines

from transformers import AutoModelForCausalLM, BitsAndBytesConfig
from peft import get_peft_model, LoraConfig, prepare_model_for_kbit_training

# 4-bit quantization config
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",           # NormalFloat4 (best for LLMs)
    bnb_4bit_compute_dtype="bfloat16",   # Compute in bf16
    bnb_4bit_use_double_quant=True       # Nested quantization
)

# Load quantized model
model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-3.1-70B",
    quantization_config=bnb_config,
    device_map="auto"
)

# Prepare for training (enables gradient checkpointing)
model = prepare_model_for_kbit_training(model)

# LoRA config for QLoRA
lora_config = LoraConfig(
    r=64,                              # Higher rank for 70B
    lora_alpha=128,
    lora_dropout=0.1,
    target_modules=["q_proj", "v_proj", "k_proj", "o_proj", "gate_proj", "up_proj", "down_proj"],
    bias="none",
    task_type="CAUSAL_LM"
)

model = get_peft_model(model, lora_config)
# 70B model now fits on single 24GB GPU!

LoRA parameter selection

A lookup table. Do not read it all; find the row that applies to you.

Rank (r) - capacity vs efficiency

Rank	Trainable Params	Memory	Quality	Use Case
4	~3M	Minimal	Lower	Simple tasks, prototyping
8	~7M	Low	Good	Recommended starting point
16	~14M	Medium	Better	General fine-tuning
32	~27M	Higher	High	Complex tasks
64	~54M	High	Highest	Domain adaptation, 70B models

Alpha (loraalpha) - scaling factor

Python4 lines

# Rule of thumb: alpha = 2 * rank
LoraConfig(r=16, lora_alpha=32)  # Standard
LoraConfig(r=16, lora_alpha=16)  # Conservative (lower learning rate effect)
LoraConfig(r=16, lora_alpha=64)  # Aggressive (higher learning rate effect)

Target modules by architecture

Python14 lines

# Llama / Mistral / Qwen
target_modules = ["q_proj", "v_proj", "k_proj", "o_proj", "gate_proj", "up_proj", "down_proj"]

# GPT-2 / GPT-Neo
target_modules = ["c_attn", "c_proj", "c_fc"]

# Falcon
target_modules = ["query_key_value", "dense", "dense_h_to_4h", "dense_4h_to_h"]

# BLOOM
target_modules = ["query_key_value", "dense", "dense_h_to_4h", "dense_4h_to_h"]

# Auto-detect all linear layers
target_modules = "all-linear"  # PEFT 0.6.0+

Loading and merging adapters

Explains the idea itself. Read it slowly; the later sections build on it.

Load trained adapter

Python12 lines

from peft import PeftModel, AutoPeftModelForCausalLM
from transformers import AutoModelForCausalLM

# Option 1: Load with PeftModel
base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B")
model = PeftModel.from_pretrained(base_model, "./lora-llama-adapter")

# Option 2: Load directly (recommended)
model = AutoPeftModelForCausalLM.from_pretrained(
    "./lora-llama-adapter",
    device_map="auto"
)

Merge adapter into base model

Python9 lines

# Merge for deployment (no adapter overhead)
merged_model = model.merge_and_unload()

# Save merged model
merged_model.save_pretrained("./llama-merged")
tokenizer.save_pretrained("./llama-merged")

# Push to Hub
merged_model.push_to_hub("username/llama-finetuned")

Multi-adapter serving

Python19 lines

from peft import PeftModel

# Load base with first adapter
model = AutoPeftModelForCausalLM.from_pretrained("./adapter-task1")

# Load additional adapters
model.load_adapter("./adapter-task2", adapter_name="task2")
model.load_adapter("./adapter-task3", adapter_name="task3")

# Switch between adapters at runtime
model.set_adapter("task1")  # Use task1 adapter
output1 = model.generate(**inputs)

model.set_adapter("task2")  # Switch to task2
output2 = model.generate(**inputs)

# Disable adapters (use base model)
with model.disable_adapter():
    base_output = model.generate(**inputs)

PEFT methods comparison

A lookup table. Do not read it all; find the row that applies to you.

Method	Trainable %	Memory	Speed	Best For
LoRA	0.1-1%	Low	Fast	General fine-tuning
QLoRA	0.1-1%	Very Low	Medium	Memory-constrained
AdaLoRA	0.1-1%	Low	Medium	Automatic rank selection
IA3	0.01%	Minimal	Fastest	Few-shot adaptation
Prefix Tuning	0.1%	Low	Medium	Generation control
Prompt Tuning	0.001%	Minimal	Fast	Simple task adaptation
P-Tuning v2	0.1%	Low	Medium	NLU tasks

IA3 (minimal parameters)

Python8 lines

from peft import IA3Config

ia3_config = IA3Config(
    target_modules=["q_proj", "v_proj", "k_proj", "down_proj"],
    feedforward_modules=["down_proj"]
)
model = get_peft_model(model, ia3_config)
# Trains only 0.01% of parameters!

Prefix Tuning

Python8 lines

from peft import PrefixTuningConfig

prefix_config = PrefixTuningConfig(
    task_type="CAUSAL_LM",
    num_virtual_tokens=20,      # Prepended tokens
    prefix_projection=True       # Use MLP projection
)
model = get_peft_model(model, prefix_config)

Integration patterns

Settings you configure once. Change one at a time so you can see what each does.

With TRL (SFTTrainer)

Python12 lines

from trl import SFTTrainer, SFTConfig
from peft import LoraConfig

lora_config = LoraConfig(r=16, lora_alpha=32, target_modules="all-linear")

trainer = SFTTrainer(
    model=model,
    args=SFTConfig(output_dir="./output", max_seq_length=512),
    train_dataset=dataset,
    peft_config=lora_config,  # Pass LoRA config directly
)
trainer.train()

With Axolotl (YAML config)

YAML11 lines

# axolotl config.yaml
adapter: lora
lora_r: 16
lora_alpha: 32
lora_dropout: 0.05
lora_target_modules:
  - q_proj
  - v_proj
  - k_proj
  - o_proj
lora_target_linear: true  # Target all linear layers

With vLLM (inference)

Python11 lines

from vllm import LLM
from vllm.lora.request import LoRARequest

# Load base model with LoRA support
llm = LLM(model="meta-llama/Llama-3.1-8B", enable_lora=True)

# Serve with adapter
outputs = llm.generate(
    prompts,
    lora_request=LoRARequest("adapter1", 1, "./lora-adapter")
)

Performance benchmarks

A lookup table. Do not read it all; find the row that applies to you.

Memory usage (Llama 3.1 8B)

Method	GPU Memory	Trainable Params
Full fine-tuning	60+ GB	8B (100%)
LoRA r=16	18 GB	14M (0.17%)
QLoRA r=16	6 GB	14M (0.17%)
IA3	16 GB	800K (0.01%)

Training speed (A100 80GB)

Method	Tokens/sec	vs Full FT
Full FT	2,500	1x
LoRA	3,200	1.3x
QLoRA	2,100	0.84x

Quality (MMLU benchmark)

Model	Full FT	LoRA	QLoRA
Llama 2-7B	45.3	44.8	44.1
Llama 2-13B	54.8	54.2	53.5

Common issues

Explains the idea itself. Read it slowly; the later sections build on it.

CUDA OOM during training

Python12 lines

# Solution 1: Enable gradient checkpointing
model.gradient_checkpointing_enable()

# Solution 2: Reduce batch size + increase accumulation
TrainingArguments(
    per_device_train_batch_size=1,
    gradient_accumulation_steps=16
)

# Solution 3: Use QLoRA
from transformers import BitsAndBytesConfig
bnb_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4")

Adapter not applying

Python8 lines

# Verify adapter is active
print(model.active_adapters)  # Should show adapter name

# Check trainable parameters
model.print_trainable_parameters()

# Ensure model in training mode
model.train()

Quality degradation

Python11 lines

# Increase rank
LoraConfig(r=32, lora_alpha=64)

# Target more modules
target_modules = "all-linear"

# Use more training data and epochs
TrainingArguments(num_train_epochs=5)

# Lower learning rate
TrainingArguments(learning_rate=1e-4)

Best practices

Explains the idea itself. Read it slowly; the later sections build on it.

Start with r=8-16, increase if quality insufficient
*Use alpha = 2 rank** as starting point
Target attention + MLP layers for best quality/efficiency
Enable gradient checkpointing for memory savings
Save adapters frequently (small files, easy rollback)
Evaluate on held-out data before merging
Use QLoRA for 70B+ models on consumer hardware

References

Explains the idea itself. Read it slowly; the later sections build on it.

Advanced Usage ↗ - DoRA, LoftQ, rank stabilization, custom modules
Troubleshooting ↗ - Common errors, debugging, optimization

Resources

Explains the idea itself. Read it slowly; the later sections build on it.

GitHub: https://github.com/huggingface/peft
Docs: https://huggingface.co/docs/peft
LoRA Paper: arXiv:2106.09685
QLoRA Paper: arXiv:2305.14314
Models: https://huggingface.co/models?library=peft