@inproceedings{ramirez-etal-2024-cache,
title = "Cache {\&} Distil: Optimising {API} Calls to Large Language Models",
author = "Ram{\'\i}rez, Guillem and
Lindemann, Matthias and
Birch, Alexandra and
Titov, Ivan",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Findings of the Association for Computational Linguistics ACL 2024",
month = aug,
year = "2024",
address = "Bangkok, Thailand and virtual meeting",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-acl.704",
pages = "11838--11853",
abstract = "Large-scale deployment of generative AI tools often depends on costly API calls to a Large Language Model (LLM) to fulfil user queries, a process that also exposes the request stream to external providers. To curtail the frequency of these calls, one can employ a local smaller language model -a student- which is continuously trained on the responses of the LLM. This student gradually gains proficiency in independently handling an increasing number of user requests, a process we term neural caching. The crucial element in neural caching is a policy that decides which requests should be processed by the student alone and which should be redirected to the LLM, subsequently aiding the student{'}s learning. In this study, we focus on classification tasks, and we consider a range of classic Active Learning-based selection criteria as the policy. Our experiments suggest that Margin Sampling and Query by Committee bring consistent benefits over other policies and baselines across tasks and budgets.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ramirez-etal-2024-cache">
<titleInfo>
<title>Cache & Distil: Optimising API Calls to Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Guillem</namePart>
<namePart type="family">Ramírez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Matthias</namePart>
<namePart type="family">Lindemann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexandra</namePart>
<namePart type="family">Birch</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ivan</namePart>
<namePart type="family">Titov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics ACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lun-Wei</namePart>
<namePart type="family">Ku</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Martins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Srikumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand and virtual meeting</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Large-scale deployment of generative AI tools often depends on costly API calls to a Large Language Model (LLM) to fulfil user queries, a process that also exposes the request stream to external providers. To curtail the frequency of these calls, one can employ a local smaller language model -a student- which is continuously trained on the responses of the LLM. This student gradually gains proficiency in independently handling an increasing number of user requests, a process we term neural caching. The crucial element in neural caching is a policy that decides which requests should be processed by the student alone and which should be redirected to the LLM, subsequently aiding the student’s learning. In this study, we focus on classification tasks, and we consider a range of classic Active Learning-based selection criteria as the policy. Our experiments suggest that Margin Sampling and Query by Committee bring consistent benefits over other policies and baselines across tasks and budgets.</abstract>
<identifier type="citekey">ramirez-etal-2024-cache</identifier>
<location>
<url>https://aclanthology.org/2024.findings-acl.704</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>11838</start>
<end>11853</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Cache & Distil: Optimising API Calls to Large Language Models
%A Ramírez, Guillem
%A Lindemann, Matthias
%A Birch, Alexandra
%A Titov, Ivan
%Y Ku, Lun-Wei
%Y Martins, Andre
%Y Srikumar, Vivek
%S Findings of the Association for Computational Linguistics ACL 2024
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand and virtual meeting
%F ramirez-etal-2024-cache
%X Large-scale deployment of generative AI tools often depends on costly API calls to a Large Language Model (LLM) to fulfil user queries, a process that also exposes the request stream to external providers. To curtail the frequency of these calls, one can employ a local smaller language model -a student- which is continuously trained on the responses of the LLM. This student gradually gains proficiency in independently handling an increasing number of user requests, a process we term neural caching. The crucial element in neural caching is a policy that decides which requests should be processed by the student alone and which should be redirected to the LLM, subsequently aiding the student’s learning. In this study, we focus on classification tasks, and we consider a range of classic Active Learning-based selection criteria as the policy. Our experiments suggest that Margin Sampling and Query by Committee bring consistent benefits over other policies and baselines across tasks and budgets.
%U https://aclanthology.org/2024.findings-acl.704
%P 11838-11853
Markdown (Informal)
[Cache & Distil: Optimising API Calls to Large Language Models](https://aclanthology.org/2024.findings-acl.704) (Ramírez et al., Findings 2024)
ACL
- Guillem Ramírez, Matthias Lindemann, Alexandra Birch, and Ivan Titov. 2024. Cache & Distil: Optimising API Calls to Large Language Models. In Findings of the Association for Computational Linguistics ACL 2024, pages 11838–11853, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.