Group Relative Policy Optimization for Speech Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shivakumar, Prashanth Gurunath, Gu, Yile, Gandhe, Ankur, Bulyko, Ivan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915474566021120
author Shivakumar, Prashanth Gurunath
Gu, Yile
Gandhe, Ankur
Bulyko, Ivan
author_facet Shivakumar, Prashanth Gurunath
Gu, Yile
Gandhe, Ankur
Bulyko, Ivan
contents Speech Recognition has seen a dramatic shift towards adopting Large Language Models (LLMs). This shift is partly driven by good scalability properties demonstrated by LLMs, ability to leverage large amounts of labelled, unlabelled speech and text data, streaming capabilities with auto-regressive framework and multi-tasking with instruction following characteristics of LLMs. However, simple next-token prediction objective, typically employed with LLMs, have certain limitations in performance and challenges with hallucinations. In this paper, we propose application of Group Relative Policy Optimization (GRPO) to enable reinforcement learning from human feedback for automatic speech recognition (ASR). We design simple rule based reward functions to guide the policy updates. We demonstrate significant improvements in word error rate (upto 18.4% relative), reduction in hallucinations, increased robustness on out-of-domain datasets and effectiveness in domain adaptation.
format Preprint
id arxiv_https___arxiv_org_abs_2509_01939
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Group Relative Policy Optimization for Speech Recognition
Shivakumar, Prashanth Gurunath
Gu, Yile
Gandhe, Ankur
Bulyko, Ivan
Audio and Speech Processing
Speech Recognition has seen a dramatic shift towards adopting Large Language Models (LLMs). This shift is partly driven by good scalability properties demonstrated by LLMs, ability to leverage large amounts of labelled, unlabelled speech and text data, streaming capabilities with auto-regressive framework and multi-tasking with instruction following characteristics of LLMs. However, simple next-token prediction objective, typically employed with LLMs, have certain limitations in performance and challenges with hallucinations. In this paper, we propose application of Group Relative Policy Optimization (GRPO) to enable reinforcement learning from human feedback for automatic speech recognition (ASR). We design simple rule based reward functions to guide the policy updates. We demonstrate significant improvements in word error rate (upto 18.4% relative), reduction in hallucinations, increased robustness on out-of-domain datasets and effectiveness in domain adaptation.
title Group Relative Policy Optimization for Speech Recognition
topic Audio and Speech Processing
url https://arxiv.org/abs/2509.01939