Metraipolis › archive › dc_b684d60bedf9
Understanding LoRA Rank Trade-offs in Diffusion Model Fine-Tuning — arXiv:2609.10656
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.10656
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T05:23:37.175Z
- SHA-256 of the archived text
b4d49dd66123d82bf79a1a686f377567d080f08542a095e541e70dc1337d1c71- Stable URL of this page
https://metraipolis.com/documento/dc_b684d60bedf9
How to cite
Understanding LoRA Rank Trade-offs in Diffusion Model Fine-Tuning — arXiv:2609.10656 — arXiv (en), https://arxiv.org/abs/2609.10656 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b684d60bedf9 (Metraipolis archive dc_b684d60bedf9, sha256 b4d49dd66123d82bf79a1a686f377567d080f08542a095e541e70dc1337d1c71)
BibTeX
@misc{dcb684d60bedf9,
title = {Understanding LoRA Rank Trade-offs in Diffusion Model Fine-Tuning — arXiv:2609.10656},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.10656},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b684d60bedf9; sha256 b4d49dd66123d82bf79a1a686f377567d080f08542a095e541e70dc1337d1c71},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.10656. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Selecting LoRA rank for diffusion fine-tuning requires balancing quality and compute cost. We present a controlled study on CIFAR-10 using a DDPM U-Net with ranks {2,4,8,16,32}, fixed optimization settings, and a reproducible local-folder pytorch-fid protocol. We report FID, trainable parameters, runtime, and GPU memory, then validate trends with extended-budget DDPM runs (20 epochs; ranks 4/8/16) and a Tiny DiT backbone (10 epochs; ranks 4/8/16). Results show moderate ranks are most efficient: rank 4 achieves the best DDPM FID (124.1380), rank 8 is close (124.2136), and higher ranks provide limited gains despite larger adaptation cost. These findings support small-to-moderate ranks as practical defaults under fixed training budgets.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-11 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.10656 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T05:17:51.948Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).