pith. sign in

arxiv: 1909.11299 · v2 · pith:YOZB7G5Vnew · submitted 2019-09-25 · 💻 cs.LG · stat.ML

Mixout: Effective Regularization to Finetune Large-scale Pretrained Language Models

classification 💻 cs.LG stat.ML
keywords finetuninglanguagemixoutpretraineddownstreammodelmodelsregularization
0
0 comments X
read the original abstract

In natural language processing, it has been observed recently that generalization could be greatly improved by finetuning a large-scale language model pretrained on a large unlabeled corpus. Despite its recent success and wide adoption, finetuning a large pretrained language model on a downstream task is prone to degenerate performance when there are only a small number of training instances available. In this paper, we introduce a new regularization technique, to which we refer as "mixout", motivated by dropout. Mixout stochastically mixes the parameters of two models. We show that our mixout technique regularizes learning to minimize the deviation from one of the two models and that the strength of regularization adapts along the optimization trajectory. We empirically evaluate the proposed mixout and its variants on finetuning a pretrained language model on downstream tasks. More specifically, we demonstrate that the stability of finetuning and the average accuracy greatly increase when we use the proposed approach to regularize finetuning of BERT on downstream tasks in GLUE.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. A Rank Stabilization Scaling Factor for Fine-Tuning with LoRA

    cs.CL 2023-11 unverdicted novelty 7.0

    LoRA adapters should be scaled by 1/sqrt(rank) rather than 1/rank to stabilize learning and enable effective use of higher ranks during fine-tuning of large language models.

  2. LLM+P: Empowering Large Language Models with Optimal Planning Proficiency

    cs.AI 2023-04 accept novelty 7.0

    LLM+P lets LLMs solve planning problems optimally by converting them to PDDL for classical planners and back to natural language.

  3. Robust Policy Optimization to Prevent Catastrophic Forgetting

    cs.LG 2026-02 unverdicted novelty 6.0

    FRPO applies a max-min robust optimization over KL-bounded policy neighborhoods during RLHF to reduce catastrophic forgetting of safety and accuracy under subsequent SFT or RL fine-tuning.

  4. Should We Still Pretrain Encoders with Masked Language Modeling?

    cs.CL 2025-07 accept novelty 6.0

    Controlled ablations of 38 models find MLM superior to CLM on representation benchmarks while CLM offers better data efficiency and stability; a biphasic CLM-then-MLM schedule is optimal under fixed compute and improv...