# AGI-Edgerunners/LLM-Adapters

Code for our EMNLP 2023 Paper: "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models"

Repository: https://github.com/AGI-Edgerunners/LLM-Adapters
Canonical: https://ross.abutalabs.com/products/llm-adapters
Homepage: https://arxiv.org/abs/2304.01933
Language: Python
License: Apache-2.0
License Family: permissive
Topics: adapters, fine-tuning, large-language-models, parameter-efficient
Last push: 2024-03-10T08:20:15+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 89
- inputs: {"age_days": 1253, "days_push": 906, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1233, forks 115 (observed 2026-08-28T04:04:04.598691+00:00)

## What it is
LLM-Adapters is a Python framework extending HuggingFace's PEFT library that integrates various adapter types (LoRA, series/parallel adapters, prefix/prompt tuning) into open-access LLMs like LLaMA, BLOOM, and GPT-J for parameter-efficient fine-tuning. It accompanies an EMNLP 2023 paper and includes benchmark datasets such as math10k and commonsense170k.

## Use cases
- fine-tune llama with lora adapters
- parameter-efficient fine-tuning of large language models
- compare adapter types like bottleneck vs parallel adapters on llms
- reproduce emnlp 2023 llm-adapters paper experiments
- train adapters on commonsense170k dataset
- apply prefix tuning or p-tuning to open-access llms

## When to choose
- you want a unified framework to experiment with multiple PEFT adapter methods on LLaMA, BLOOM, GPT-J, or OPT
- you need the paper's benchmark datasets and adapter checkpoints for research reproduction
- you want to study adapter placement and hyperparameter effects on fine-tuning performance

## When to avoid
- you need actively maintained tooling with support for the latest LLMs - the repo has seen limited updates since 2024
- you want full fine-tuning rather than parameter-efficient methods
- you need production-grade training infrastructure rather than research code

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-training, machine-learning, sdk
- domain: large-language-models, machine-learning, deep-learning
- platform: python
- tags: peft, lora, adapters, fine-tuning, prompt-tuning, huggingface, research-code, llama, natural-language-processing, gpu, linux

## Member repositories
- AGI-Edgerunners/LLM-Adapters (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:04.598691+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:22:12.680865+00:00, confidence not recorded.
  - readme: https://github.com/AGI-Edgerunners/LLM-Adapters (fetched 2026-08-28T04:04:04.598691+00:00, sha 854425ad888d)
  - homepage: https://arxiv.org/abs/2304.01933 (fetched 2026-08-29T12:22:01.962058+00:00, sha 90cdcbf30a03)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:22:01.971349+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:22:01.975397+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:22:01.977519+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:22:01.973196+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.