Federated Adaptive Prompt Tuning for Multi-Domain Collaborative Learning

Authors: Shangchao Su, Mingzhao Yang, Bin Li, Xiangyang Xue

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We perform extensive experiments on two multi-domain image classification datasets across two different settings supervised and unsupervised. The results show that Fed APT can achieve better performance with less than 10% of the number of parameters of the fully trained model, and the global model can perform well in diverse client domains simultaneously.
Researcher Affiliation Academia Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science, Fudan University {scsu20, mzyang20, libin, xyxue}@fudan.edu.cn
Pseudocode No The paper describes the algorithm steps in text and flow diagrams but does not provide a formal pseudocode or algorithm block.
Open Source Code No The paper does not provide any statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We adopt two datasets, Office-Caltech10 (Gong et al. 2012) and Domain Net (Peng et al. 2019a).
Dataset Splits No The paper describes how datasets are split across clients and domains (e.g., 'we use each domain as a client', 'split each domain in Domain Net into five clients'), but it does not specify explicit train/validation/test splits (e.g., percentages or counts) for the overall datasets or individual client datasets.
Hardware Specification Yes All experiments are completed with one Ge Force RTX 3090 GPU.
Software Dependencies No The paper mentions 'We use Py Torch to implement all methods' but does not specify the version number of PyTorch or any other software dependencies.
Experiment Setup Yes We set Office-Caltech10 with a learning rate of 0.001 and batch size of 32, and Domain Net with a learning rate of 0.01 and batch size of 256. The global communication round Tg is set to 50, and the local training epoch Tl is set to 1. The length of prompts s is 16.