On Surgical Fine-tuning for Language Encoders

Abhilasha Lodha, Gayatri Vyankatesh Belapurkar, Saloni Chalkapurkar, Yuanming Tao, Reshmi Ghosh, Samyadeep Basu, Dmitrii M Petrov, Soundararajan Srinivasan

Submission Type: Regular Short Paper

Submission Track: Efficient Methods for NLP

Submission Track 2: Language Modeling and Analysis of Language Models

Keywords: Efficient Fine Tuning, Language Models, Language Encoders, Linguistics, Optimization, Distributional Shifts, Temporal Shifts

TL;DR: Using Fisher Information Matrix diagonal score we perform efficient surgical fine-tuning for LM tasks

Abstract: Fine-tuning all the layers of a pre-trained neural language encoder (either using all the parameters or using parameter-efficient methods) is often the de-facto way of adapting it to a new task. We show evidence that for different downstream language tasks, fine-tuning only a subset of layers is sufficient to obtain performance that is close to and often better than fine-tuning all the layers in the language encoder. We propose an efficient metric based on the diagonal of the Fisher information matrix (FIM score), to select the candidate layers for selective fine-tuning. We show, empirically on GLUE and SuperGLUE tasks and across distinct language encoders, that this metric can effectively select layers leading to a strong downstream performance. Our work highlights that task-specific information corresponding to a given downstream task is often localized within a few layers, and tuning only those is sufficient for strong performance. Additionally, we demonstrate the robustness of the FIM score to rank layers in a manner that remains constant during the optimization process.

Submission Number: 4258