PARP: Prune, Adjust and Re-Prune for Self-Supervised Speech Recognition

Abstract

Self-supervised speech representation learning (speech SSL) has demonstrated the benefit of scale in learning rich representations for Automatic Speech Recognition (ASR) with limited paired data, such as wav2vec 2.0. We investigate the existence of sparse subnetworks in pre-trained speech SSL models that achieve even better low-resource ASR results. However, directly applying widely adopted pruning methods such as the Lottery Ticket Hypothesis (LTH) is suboptimal in the computational cost needed. Moreover, we show that the discovered subnetworks yield minimal performance gain compared to the original dense network. We present Prune-Adjust-Re-Prune (PARP), which discovers and finetunes subnetworks for much better performance, while only requiring a single downstream ASR finetuning run. PARP is inspired by our surprising observation that subnetworks pruned for pre-training tasks need merely a slight adjustment to achieve a sizeable performance boost in downstream ASR tasks. Extensive experiments on low-resource ASR verify (1) sparse subnetworks exist in mono-lingual/multi-lingual pre-trained speech SSL, and (2) the computational advantage and performance gain of PARP over baseline pruning methods. In particular, on the 10min Librispeech split without LM decoding, PARP discovers subnetworks from wav2vec 2.0 with an absolute 10.9%/12.6% WER decrease compared to the full model. We further demonstrate the effectiveness of PARP via: cross-lingual pruning without any phone recognition degradation, the discovery of a multi-lingual subnetwork for 10 spoken languages in 1 finetuning run, and its applicability to pre-trained BERT/XLNet for natural language tasks.

References in corpus (13)

• The State of Sparsity in Deep Neural Networks
• Conformer: Convolution-augmented Transformer for Speech Recognition
• Comparing Rewinding and Fine-tuning in Neural Network Pruning
• ContextNet: Improving Convolutional Neural Networks for Automatic Speech Recognition with Global Context
• DeCoAR 2.0: Deep Contextualized Acoustic Representations with Vector Quantization
• The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models
• Contrastive Predictive Coding Based Feature for Automatic Speaker Verification
• W2v-BERT: Combining Contrastive Learning and Masked Language Modeling for Self-Supervised Speech Pre-Training
• Simple and Effective Zero-shot Cross-lingual Phoneme Recognition
• UniSpeech at scale: An Empirical Study of Pre-training Method on Large-Scale Speech Recognition Dataset
• Mandarin-English Code-switching Speech Recognition with Self-supervised Speech Representation Models
• Speech Representation Learning Through Self-supervised Pretraining And Multi-task Finetuning
• Large-Scale Pre-Training of End-to-End Multi-Talker ASR for Meeting Transcription with Single Distant Microphone