{
  "abstract": "Background: Protein language models (PLMs) are increasingly used for sequence generation and property prediction, but their size makes fine-tuning and deployment expensive. The effects of quantization and parameter efficient fine-tuning on performance, representations and generation remain insufficiently characterized. Results: We evaluated 4-bit quantization and low-rank adapter fine-tuning (QLoRA) across ESM-2, ESMC, ProtBERT, ProtT5, Ankh, Ankh3 and Profluent-E1. Across protein prediction tasks, many model-task pairs retained more than 90% of full fine-tuning performance. Peak GPU memory savings approached 90% for the largest models, although performance and efficiency varied by model, dataset and training configuration. QLoRA often preserved early-layer representations while inducing task-specific adaptations in middle and late layers, resembling full fine-tuning with smaller representational changes. Training speed and power effects were more varied. For unconditional generation with ProLLaMA, ProtGPT2, ProGen2, ProteinGLM and ESM3, 4-bit quantization largely preserved predicted structural and sequence-level properties, but token-level analysis revealed model-dependent shifts in autoregressive output distributions. Conclusion: QLoRA and 4-bit quantization reduce PLM computational requirements, particularly GPU memory usage. Our results support QLoRA as a first-pass strategy for memory limited adaptation, reserving full fine-tuning for challenging tasks, unstable architectures or low validation recovery. For generative PLMs, sequence-level and structural metrics should be complemented with distributional analysis, since downstream predictions alone may miss quantization-induced shifts. These approaches can broaden access to large-scale protein modelling while requiring model- and task-specific validation.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2610.00665",
  "author_total": 11,
  "authors": [
    {
      "name": "Ilan Yaniv Zeisler"
    },
    {
      "name": "Sebastian Clancy"
    },
    {
      "name": "Pouriya Bayat"
    },
    {
      "name": "Saaim Raad"
    },
    {
      "name": "Ivan Kraskov"
    },
    {
      "name": "Matthew Xie"
    },
    {
      "name": "Vivian White"
    },
    {
      "name": "Spencer Perkins"
    },
    {
      "name": "Serena Singh"
    },
    {
      "name": "Sepehr Bayat"
    },
    {
      "name": "Keith Pardee"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/437hev38yf64x.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.00665",
  "fulltext_url": null,
  "id": "437hev38yf64x",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.00665",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "q-bio.QM"
    ]
  ],
  "openalex": null,
  "pdf_url": "https://arxiv.org/pdf/2610.00665",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-01",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/437hev38yf64x.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:26:08Z",
      "native_id": "2610.00665",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.00665"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "arxiv:q-bio.QM",
      "subfield": "bioinformatics"
    }
  ],
  "title": "Analysis of Quantized and Efficiently Adapted Protein Language Models",
  "title_text": "Analysis of Quantized and Efficiently Adapted Protein Language Models",
  "type": "preprint",
  "updated_at": "2026-10-30T12:26:08Z",
  "url": "https://research.draconitic.com/paper/437hev38yf64x",
  "venue": "arXiv",
  "versions": []
}