PFedQFT: Personalized Federated Quantum-Inspired Fine-Tuning for Medical Large Language Models

Qihan Huang; Lingxia Zhang; Guye Zhang; Zizhu Wang; Xiaowei Mao

cs.AIOct 19, 2026

PFedQFT: Personalized Federated Quantum-Inspired Fine-Tuning for Medical Large Language Models

Authors: Qihan Huang, Lingxia Zhang, Guye Zhang, Zizhu Wang, Xiaowei Mao

Organizations: Institute of Fundamental and Frontier Sciences, University of Electronic Science and Technology of China, Chengdu, Sichuan, China · Key Laboratory of Quantum Physics and Photonic Quantum Information, Ministry of Education, University of Electronic Science and Technology of China, Chengdu, Sichuan, China · Sichuan Provincial Key Laboratory for Human Disease Gene Study and the Center for Medical Genetics, Department of Laboratory Medicine, Sichuan Academy of Medical Sciences and Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China · Sichuan Quantum Science and Technology Institute, Chengdu, Sichuan, China

Abstract

Medical data remain fragmented across institutional silos because of privacy, governance, and ownership constraints. Federated learning enables medical institutions to adapt large language models (LLMs) collaboratively without transferring raw records. However, independently parameterized low-rank adapters incur repeated parameter and communication costs across Transformer projections, while uniform aggregation may limit adaptation to heterogeneous institutional data. We propose PFedQFT, a quantum-inspired framework for personalized federated fine-tuning with a compact shared adapter. Specifically, PFedQFT constructs reusable orthonormal adaptation bases from compact skew-symmetric generators using a truncated matrix-exponential approximation followed by QR orthonormalization. Compatible projections reuse these bases, while a globally shared full core preserves the flexibility of each projection. Each client further maintains a lightweight private residual operator, whose directions are learned globally and whose coefficients remain local. Experiments on seven public medical tasks and multiple LLM backbones show substantial reductions in trainable parameters and per-round adapter upload volume, with competitive predictive performance on most tasks.

Figures & tables

Fig. 1: Challenges in FFT of medical LLMs: (a) institutional data silos, (b) high communication costs from repeated model update exchanges, and (c) heterogeneous data across institutions.

Fig. 2: Overall architecture of PFedQFT. The framework comprises a global server that coordinates sample-weighted aggregation and multiple clients that construct reusable adaptation subspaces while retaining private core residuals for personalized local training. (a) At communication round $t$ , the server selects the participating clients $\mathcal{S}_{t}$ and broadcasts the shared parameters $\Theta^{t}$ . (b) Each selected client transforms the generators $\mathbf{B}_{g}^{\mathrm{in}}$ and $\mathbf{B}_{g}^{\mathrm{out}}$ associated with sharing group $g$ into the orthonormal bases $\mathbf{V}_{g}$ and $\mathbf{U}_{g}$ through Taylor mapping and thin QR. (c) For each adapted projection $\ell$ , client $k$ constructs the private residual operator $\mathbf{D}_{k,\ell}=\mathbf{R}_{\ell}^{\mathrm{out}}\operatorname{diag}(\mathbf{p}_{k,\ell})(\mathbf{R}_{\ell}^{\mathrm{in}})^{\top}$ and the personalized subspace operator $\mathbf{C}_{k,\ell}=\mathbf{G}_{\ell}+\mathbf{D}_{k,\ell}$ . (d) Local optimization updates the shared parameter copy $\Theta_{k}$ and the private coefficients $\phi_{k}$ using $\mathcal{D}_{k}$ . (e) Client $k$ uploads only $(\Delta\Theta_{k}^{t},n_{k})$ , while $\mathcal{D}_{k}$ and $\mathbf{p}_{k,\ell}$ remain local. (f) The server performs sample-weighted aggregation to obtain the shared parameters $\Theta^{t+1}$ for the next communication round.

Fig. 3: Overall performance comparison. The Llama-3.2-3B-Instruct backbone is frozen, and only adapter parameters are optimized with five participating clients per communication round. The LoRA baselines use local rank-16 adapters, whereas PFedQFT uses task-specific adaptation configurations with globally shared parameters and residual coefficients retained privately by each client. Results are evaluated on the complete validation sets using Accuracy for medical question answering and Soft F1 for diagnosis normalization and information extraction.

Fig. 4: Performance across backbone models. All backbones are frozen, and only adapter parameters are optimized with one local epoch per communication round, an effective batch size of 16, and a maximum sequence length of 768. FLoRA uses rank-16 adapters, whereas PFedQFT uses an adaptation rank of 8 and an intrinsic rank of 4 with residual coefficients retained privately by each client. Accuracy is evaluated on the complete validation splits, and

\Delta

denotes the accuracy of PFedQFT minus that of FLoRA in percentage points.

Non-IID Degree	MedQA-MCMLE		CHIP-CDN		CMeEE-V2
Non-IID Degree	FLoRA	PFedQFT	FLoRA	PFedQFT	FLoRA	PFedQFT
Light	81.83	81.49	78.32	76.52	77.79	76.74
Moderate	81.72	81.46	79.95	76.57	77.12	76.35
Severe	81.15	81.14	77.96	76.08	77.19	76.56

TABLE I: Performance under varying degrees of data heterogeneity.

Method	#TP (M)	TP (%)	Shared (MB)	Private (MB)	Upload (MB)	Time
FLoRA †	24.3139	0.7568	97.2554	–	97.2554	0 h 45 min
FedAdam-LoRA	24.3139	0.7568	97.2554	–	97.2554	0 h 37 min
FFA-LoRA	12.3863	0.3855	49.5452	–	49.5452	1 h 06 min
FedEx-LoRA ‡	24.3139	0.7568	97.2554	–	97.2554	0 h 52 min
PFedQFT	1.6524	0.0514	6.5981	0.011648	6.5981	6 h 23 min

TABLE II: Average trainable parameter count, adapter storage, communication cost, and runtime across the seven tasks.

Fig. 5: Personalization mechanism ablation on the 196-example CHIP-CDN target-client validation set. Global disables private adaptation; PFedQFT-D trains the private branch on local examples whose diagnosis concepts do not overlap with the validation set; and PFedQFT uses local support covering the target concepts. Moderate and Strong denote private-branch update settings

(\eta_{p},s_{p},\lambda_{p})=(0.0015,0.12,5\!\times\!10^{-5})

and

(0.0020,0.16,2\!\times\!10^{-5})

, respectively. Shared parameters remain frozen in all personalized runs. Each distribution is obtained from 3,000 bootstrap resamples of the validation examples rather than independent training runs; dashed lines indicate Soft F1 on the complete validation set.

Federated Fine-Tuning

Federated fine-tuning adapts pretrained models across decentralized data holders without pooling their private data.

See all 115 papers

FedFit: Federated Fine-Tuning of LLMs via Vector-Bank Parameterization and Quantization
Oct 1, 2026Hang Zou, Chao Zhang, Yuzhi Yang +3Federated Learning Aggregation Low-Rank Adaptation
6G Research Center, Khalifa University, 127788 Abu Dhabi, UAE · Department of Computer Science, Central South University, 410083, Changsha, China · Université de Lorraine, CNRS, CRAN, F-54000 Nancy, France
FraQ: Efficient Coordinate-Space Recompression for Federated Low-Rank Adaptation
Aug 4, 2026Shenghui Li, Thiemo VoigtFederated Learning Aggregation Communication-Efficient Distributed Training
Uppsala University, Uppsala, Sweden · Research Institutes of Sweden, Stockholm, Sweden
FED-FSTQ: Fisher-Guided Token Quantization for Communication-Efficient Federated Fine-Tuning of LLMs on Edge Devices
Apr 28, 2026Changyu Li, Shuanghong Huang, Jiashen Liu +5Federated Fine-Tuning Federated Learning
School of Computing and Information Technology, Great Bay University, Dongguan, China · Beijing Institute of Technology, Beijing, China · University of Warwick, Coventry, U.K. +3

Oct 1, 2026J/K move · Enter open · S save

FedFit: Federated Fine-Tuning of LLMs via Vector-Bank Parameterization and Quantization

Hang Zou, Chao Zhang, Yuzhi Yang, Yu Tian+2

6G Research Center, Khalifa University, 127788 Abu Dhabi, UAE · Department of Computer Science, Central South University, 410083, Changsha, China · Université de Lorraine, CNRS, CRAN, F-54000 Nancy, France

Federated Learning Aggregation Low-Rank Adaptation Federated Fine-Tuning Federated Learning

Open paper AskPDF arXiv

Federated Learning (FL) enables privacy-preserving fine-tuning of Large Language Models (LLMs), yet the massive communication overhead remains a critical bottleneck. Furthermore, applying Low-Rank Adaptation (LoRA) in FL faces a fundamental "aggregation dilemma" between the accurate Sum-of-Products (SoP) and the communication-efficient Product-of-Sums (PoS) implementations. To tackle these challenges, we propose FedFit. First, to significantly reduce communication overhead, we introduce a disjoint shared vector-bank parameterization that reconstructs high-dimensional adapter matrices from two compact and disjoint global vector banks. Second, to address the aggregation dilemma, we devise an alternating optimization schedule. By cycling between decoupled single-bank updates (which allow for accurate aggregation) and joint updates corrected by a Residual Spectral Aggregation mechanism, we resolve the conflict between SoP and PoS. Additionally, we integrate blockwise quantization with client-side error feedback to further compress the transmitted vectors. Furthermore, we establish theoretical convergence guarantees for the proposed algorithm. Extensive experiments on Qwen2.5 models demonstrate that FedFit achieves perplexity performance comparable to standard federated LoRA methods, while providing compression ratios up to 100x higher.

Abstract

Federated Fine-Tuning

Explore similar work

FedFit: Federated Fine-Tuning of LLMs via Vector-Bank Parameterization and Quantization

FraQ: Efficient Coordinate-Space Recompression for Federated Low-Rank Adaptation

FED-FSTQ: Fisher-Guided Token Quantization for Communication-Efficient Federated Fine-Tuning of LLMs on Edge Devices