# ApproBiVT: Lead ASR Models to Generalize Better Using Approximated Bias-Variance Tradeoff Guided Early Stopping and Checkpoint Averaging

arXiv: https://arxiv.org/abs/2308.02870 · HTML: https://mlsanity.com/paper/2308.02870/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 31 references that are in the map.
- First posted on arXiv 5 Aug 2023.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Conformer: Convolution-augmented Transformer for Speech Recognition](https://mlsanity.com/paper/2005.08100/) (2020)
- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [Listen, Attend and Spell](https://mlsanity.com/paper/1508.01211/) (2015)
- [AISHELL-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baseline](https://mlsanity.com/paper/1709.05522/) (2017)
- [Train longer, generalize better: closing the generalization gap in large batch training of neural networks](https://mlsanity.com/paper/1705.08741/) (2017)
- [AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale](https://mlsanity.com/paper/1808.10583/) (2018)
- [WeNet: Production oriented Streaming and Non-streaming End-to-End Speech Recognition Toolkit](https://mlsanity.com/paper/2102.01547/) (2021)
- [Exploring Architectures, Data and Units For Streaming End-to-End Speech Recognition with RNN-Transducer](https://mlsanity.com/paper/1801.00841/) (2018)
- [QuartzNet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutions](https://mlsanity.com/paper/1910.10261/) (2019)
- [Training Tips for the Transformer Model](https://mlsanity.com/paper/1804.00247/) (2018)
- [Jasper: An End-to-End Convolutional Neural Acoustic Model](https://mlsanity.com/paper/1904.03288/) (2019)
- [Streaming automatic speech recognition with the transformer model](https://mlsanity.com/paper/2001.02674/) (2020)
- [Squeezeformer: An Efficient Transformer for Automatic Speech Recognition](https://mlsanity.com/paper/2206.00888/) (2022)
- [Unified Streaming and Non-streaming Two-pass End-to-end Model for Speech Recognition](https://mlsanity.com/paper/2012.05481/) (2020)
- [U2++: Unified Two-pass Bidirectional End-to-end Model for Speech Recognition](https://mlsanity.com/paper/2106.05642/) (2021)
- [Dropout as data augmentation](https://mlsanity.com/paper/1506.08700/) (2015)
- [Multistream CNN for Robust Acoustic Modeling](https://mlsanity.com/paper/2005.10470/) (2020)
- [Revisiting Checkpoint Averaging for Neural Machine Translation](https://mlsanity.com/paper/2210.11803/) (2022)
- [CUSIDE: Chunking, Simulating Future Context and Decoding for Streaming ASR](https://mlsanity.com/paper/2203.16758/) (2022)
- [Improving Mandarin Speech Recogntion with Block-augmented Transformer](https://mlsanity.com/paper/2207.11697/) (2022)
- [Shifted Chunk Encoder for Transformer Based Streaming End-to-End ASR](https://mlsanity.com/paper/2203.15206/) (2022)

## Abstract (arXiv)

The conventional recipe for Automatic Speech Recognition (ASR) models is to 1) train multiple checkpoints on a training set while relying on a validation set to prevent overfitting using early stopping and 2) average several last checkpoints or that of the lowest validation losses to obtain the final model. In this paper, we rethink and update the early stopping and checkpoint averaging from the perspective of the bias-variance tradeoff. Theoretically, the bias and variance represent the fitness and variability of a model and the tradeoff of them determines the overall generalization error. But, it's impractical to evaluate them precisely. As an alternative, we take the training loss and validation loss as proxies of bias and variance and guide the early stopping and checkpoint averaging using their tradeoff, namely an Approximated Bias-Variance Tradeoff (ApproBiVT). When evaluating with advanced ASR models, our recipe provides 2.5%-3.7% and 3.1%-4.6% CER reduction on the AISHELL-1 and AISHELL-2, respectively.