Détail du document
Identifiant
oai:arXiv.org:2410.22304
Sujet
Computer Science - Computation and... Computer Science - Machine Learnin...Auteur
Deng, Yihe Mineiro, PaulCatégorie
Computer Science
Année
2024
Date de référencement
06/11/2024
Mots clés
llm mathematical online reasoningRésumé
Mathematical reasoning is a crucial capability for Large Language Models (LLMs), yet generating detailed and accurate reasoning traces remains a significant challenge.
This paper introduces a novel approach to produce high-quality reasoning traces for LLM fine-tuning using online learning \textbf{Flows}.
Our method employs an incremental output production Flow, where component LLMs collaboratively construct solutions through iterative communication.
We train the Flow using online Direct Preference Optimization (DPO) learning with rollouts, generating DPO pairs for each training example and updating models in real-time.
We directly compare the quality of reasoning traces generated by our method with those produced through direct model inference, demonstrating the effectiveness of our approach in improving LLM performance in mathematical reasoning tasks.
;Comment: 5 pages, 4 figures, 1 table
Deng, Yihe,Mineiro, Paul, 2024, Flow-DPO: Improving LLM Mathematical Reasoning through Online Multi-Agent Learning
Clinical Relevance of Plaque Distribution for Basilar Artery Stenosis
study endovascular imaging wall basilar complications plaque postoperative artery plaques stenosis
