L3Ms -- Lagrange Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dhillon, Guneet S., Shi, Xingjian, Teh, Yee Whye, Smola, Alex
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913737219244032
author Dhillon, Guneet S.
Shi, Xingjian
Teh, Yee Whye
Smola, Alex
author_facet Dhillon, Guneet S.
Shi, Xingjian
Teh, Yee Whye
Smola, Alex
contents Supervised fine-tuning (SFT) and alignment of large language models (LLMs) are key steps in providing a good user experience. However, the concept of an appropriate alignment is inherently application-dependent, and current methods often rely on heuristic choices to drive optimization. In this work, we formulate SFT and alignment as a constrained optimization problem: the LLM is fine-tuned on a task while being required to meet application-specific requirements, without resorting to heuristics. To solve this, we propose Lagrange Large Language Models (L3Ms), which employ logarithmic barriers to enforce the constraints. This approach allows for the customization of L3Ms across diverse applications while avoiding heuristic-driven processes. We experimentally demonstrate the versatility and efficacy of L3Ms in achieving tailored alignments for various applications.
format Preprint
id arxiv_https___arxiv_org_abs_2410_21533
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle L3Ms -- Lagrange Large Language Models
Dhillon, Guneet S.
Shi, Xingjian
Teh, Yee Whye
Smola, Alex
Machine Learning
Artificial Intelligence
Computation and Language
Supervised fine-tuning (SFT) and alignment of large language models (LLMs) are key steps in providing a good user experience. However, the concept of an appropriate alignment is inherently application-dependent, and current methods often rely on heuristic choices to drive optimization. In this work, we formulate SFT and alignment as a constrained optimization problem: the LLM is fine-tuned on a task while being required to meet application-specific requirements, without resorting to heuristics. To solve this, we propose Lagrange Large Language Models (L3Ms), which employ logarithmic barriers to enforce the constraints. This approach allows for the customization of L3Ms across diverse applications while avoiding heuristic-driven processes. We experimentally demonstrate the versatility and efficacy of L3Ms in achieving tailored alignments for various applications.
title L3Ms -- Lagrange Large Language Models
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2410.21533