Shrinkml: End-to-end ASR Model Compression Using Reinforcement Learning

Abstract

End-to-end automatic speech recognition (ASR) models are increasingly large and complex to achieve the best possible accuracy. In this paper, we build an AutoML system that uses reinforcement learning (RL) to optimize the per-layer compression ratios when applied to a state-of-the-art attention based end-to-end ASR model composed of several LSTM layers. We use singular value decomposition (SVD) low-rank matrix factorization as the compression method. For our RL-based AutoML system, we focus on practical considerations such as the choice of the reward/punishment functions, the formation of an effective search space, and the creation of a representative but small data set for quick evaluation between search steps. Finally, we present accuracy results on LibriSpeech of the model compressed by our AutoML system, and we compare it to manually-compressed models. Our results show that in the absence of retraining our RL-based search is an effective and practical method to compress a productio

Stats

• citations17
• S2 citations—
• github stars0
• HF likes0
• heat score9.41
• arxiv keydudziak2019shrinkml

Related papers

• Usm-lite: Quantization And Sparsity Aware Fine-tuning For Speech Recognition With Universal Speech Models (2023)4.52
• Sequence-to-sequence ASR Optimization Via Reinforcement Learning (2017)9.41
• Accurate And Structured Pruning For Efficient Automatic Speech Recognition (2023)7.81
• Structured Pruning Of Self-supervised Pre-trained Models For Speech Recognition And Understanding (2023)11.39
• ML-LMCL: Mutual Learning And Large-margin Contrastive Learning For Improving ASR Robustness In Spoken Language Understanding (2023)0.00
• Extreme Encoder Output Frame Rate Reduction: Improving Computational Latencies Of Large End-to-end Models (2024)5.84
• Exploration Of Efficient End-to-end ASR Using Discretized Input From Self-supervised Learning (2023)12.02
• Integrating Pre-trained Speech And Language Models For End-to-end Speech Recognition (2023)0.00