Saved in:
Bibliographic Details
Main Authors: Rajani, Neel, Kiessling, Lilli, Ogaltsov, Aleksandr, Lang, Claus
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.13749
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910617352273920
author Rajani, Neel
Kiessling, Lilli
Ogaltsov, Aleksandr
Lang, Claus
author_facet Rajani, Neel
Kiessling, Lilli
Ogaltsov, Aleksandr
Lang, Claus
contents Although powerful, current cutting-edge LLMs may not fulfil the needs of highly specialised sectors. We introduce KodeXv0.1, a family of large language models that outclass GPT-4 in financial question answering. We utilise the base variants of Llama 3.1 8B and 70B and adapt them to the financial domain through a custom training regime. To this end, we collect and process a large number of publicly available financial documents such as earnings calls and business reports. These are used to generate a high-quality, synthetic dataset consisting of Context-Question-Answer triplets which closely mirror real-world financial tasks. Using the train split of this dataset, we perform RAG-aware 4bit LoRA instruction tuning runs of Llama 3.1 base variants to produce KodeX-8Bv0.1 and KodeX-70Bv0.1. We then complete extensive model evaluations using FinanceBench, FinQABench and the withheld test split of our dataset. Our results show that KodeX-8Bv0.1 is more reliable in financial contexts than cutting-edge instruct models in the same parameter regime, surpassing them by up to 9.24%. In addition, it is even capable of outperforming state-of-the-art proprietary models such as GPT-4 by up to 7.07%. KodeX-70Bv0.1 represents a further improvement upon this, exceeding GPT-4's performance on every tested benchmark.
format Preprint
id arxiv_https___arxiv_org_abs_2409_13749
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle KodeXv0.1: A Family of State-of-the-Art Financial Large Language Models
Rajani, Neel
Kiessling, Lilli
Ogaltsov, Aleksandr
Lang, Claus
Computation and Language
Artificial Intelligence
Computational Finance
I.2.7
Although powerful, current cutting-edge LLMs may not fulfil the needs of highly specialised sectors. We introduce KodeXv0.1, a family of large language models that outclass GPT-4 in financial question answering. We utilise the base variants of Llama 3.1 8B and 70B and adapt them to the financial domain through a custom training regime. To this end, we collect and process a large number of publicly available financial documents such as earnings calls and business reports. These are used to generate a high-quality, synthetic dataset consisting of Context-Question-Answer triplets which closely mirror real-world financial tasks. Using the train split of this dataset, we perform RAG-aware 4bit LoRA instruction tuning runs of Llama 3.1 base variants to produce KodeX-8Bv0.1 and KodeX-70Bv0.1. We then complete extensive model evaluations using FinanceBench, FinQABench and the withheld test split of our dataset. Our results show that KodeX-8Bv0.1 is more reliable in financial contexts than cutting-edge instruct models in the same parameter regime, surpassing them by up to 9.24%. In addition, it is even capable of outperforming state-of-the-art proprietary models such as GPT-4 by up to 7.07%. KodeX-70Bv0.1 represents a further improvement upon this, exceeding GPT-4's performance on every tested benchmark.
title KodeXv0.1: A Family of State-of-the-Art Financial Large Language Models
topic Computation and Language
Artificial Intelligence
Computational Finance
I.2.7
url https://arxiv.org/abs/2409.13749