Demystifying Domain-adaptive Post-training for Financial LLMs

TL;DR

This paper presents FINDAP, a systematic framework for domain-adaptive post-training of LLMs in finance, introducing new training recipes, datasets, and evaluation methods that significantly improve financial task performance.

Contribution

We propose FINDAP, a comprehensive approach combining capabilities definition, training recipes, curated datasets, and evaluation for effective financial domain adaptation of LLMs.

Findings

Llama-Fin achieves state-of-the-art results on financial tasks

Each post-training stage enhances specific capabilities

Preference data distillation improves instruction-following

Abstract

Domain-adaptive post-training of large language models (LLMs) has emerged as a promising approach for specialized domains such as medicine and finance. However, significant challenges remain in identifying optimal adaptation criteria and training strategies across varying data and model configurations. To address these challenges, we introduce FINDAP, a systematic and fine-grained investigation into domain-adaptive post-training of LLMs for the finance domain. Our approach consists of four key components: FinCap, which defines the core capabilities required for the target domain; FinRec, an effective training recipe that jointly optimizes continual pre-training and instruction-following, along with a novel preference data distillation method leveraging process signals from a generative reward model; FinTrain, a curated set of training datasets supporting FinRec; and FinEval, a comprehensive evaluation suite aligned with FinCap. The resulting model, Llama-Fin, achieves state-of-the-art performance across a wide range of financial tasks. Our analysis also highlights how each post-training stage contributes to distinct capabilities, uncovering specific challenges and effective solutions, providing valuable insights for domain adaptation of LLMs