A Comprehensive Analysis of Adapter Efficiency.

Nandini Mundra, Sumanth Doddapaneni, Raj Dabre, Anoop Kunchukuttan, Ratish Puduppully, Mitesh M. Khapra

Publikation: Konference artikel i Proceeding eller bog/rapport kapitelKonferencebidrag i proceedingsForskningpeer review

Abstract

Adapters have been positioned as a parameter-efficient fine-tuning (PEFT) approach, whereby a minimal number of parameters are added to the model and fine-tuned. However, adapters have not been sufficiently analyzed to understand if PEFT translates to benefits in training/deployment efficiency and maintainability/extensibility. Through extensive experiments on many adapters, tasks, and languages in supervised and cross-lingual zero-shot settings, we clearly show that for Natural Language Understanding (NLU) tasks, the parameter efficiency in adapters does not translate to efficiency gains compared to full fine-tuning of models. More precisely, adapters are relatively expensive to train and have slightly higher deployment latency. Furthermore, the maintainability/extensibility benefits of adapters can be achieved with simpler approaches like multi-task training via full fine-tuning, which also provide relatively faster training times. We, therefore, recommend that for moderately sized models for NLU tasks, practitioners should rely on full fine-tuning or multi-task training rather than using adapters. Our code is available at https://github.com/AI4Bharat/adapter-efficiency.
OriginalsprogEngelsk
TitelCODS-COMAD '24: Proceedings of the 7th Joint International Conference on Data Science & Management of Data (11th ACM IKDD CODS and 29th COMAD)
ForlagAssociation for Computing Machinery
Publikationsdato2024
Sider136-154
DOI
StatusUdgivet - 2024
Udgivet eksterntJa

Fingeraftryk

Dyk ned i forskningsemnerne om 'A Comprehensive Analysis of Adapter Efficiency.'. Sammen danner de et unikt fingeraftryk.

Citationsformater