We generalize deep self-attention distillation in MiniLM (Wang et al., 2020) by only using self-attention relation distillation for task-agnostic compression of pretrained Transformers. In particular, we define multi-head self-attention relations as scaled dot-product between the pairs of query, key, and value vectors within each self-attention module. Then we employ the above relational knowledge to train the student model. Besides its simplicity and unified principle, more favorably, there is no restriction in terms of the number of student's attention heads, while most previous work has to guarantee the same head number between teacher and student. Moreover, the fine-grained self-attention relations tend to fully exploit the interaction knowledge learned by Transformer. In addition, we thoroughly examine the layer selection strategy for teacher models, rather than just relying on the last layer as in MiniLM. We conduct extensive experiments on compressing both monolingual and multilingual pretrained models. Experimental results demonstrate that our models distilled from base-size and large-size teachers (BERT, RoBERTa and XLM-R) outperform the state-of-the-art.
In this paper, we propose Stochastic Knowledge Distillation (SKD) to obtain compact BERT-style language model dubbed SKDBERT. In each iteration, SKD samples a teacher model from a pre-defined teacher ensemble, which consists of multiple teacher models with multi-level capacities, to transfer knowledge into student model in an one-to-one manner. Sampling distribution plays an important role in SKD. We heuristically present three types of sampling distributions to assign appropriate probabilities for multi-level teacher models. SKD has two advantages: 1) it can preserve the diversities of multi-…
How to learn highly compact yet effective sentence representation? Pre-trained language models have been effective in many NLP tasks. However, these models are often huge and produce large sentence embeddings. Moreover, there is a big performance gap between large and small models. In this paper, we propose Homomorphic Projective Distillation (HPD) to learn compressed sentence embeddings. Our method augments a small Transformer encoder model with learnable projection layers to produce compact representations while mimicking a large pre-trained language model to retain the sentence representati…
Pretrained language models (PLMs) such as BERT adopt a training paradigm which first pretrain the model in general data and then finetune the model on task-specific data, and have recently achieved great success. However, PLMs are notorious for their enormous parameters and hard to be deployed on real-life applications. Knowledge distillation has been prevailing to address this problem by transferring knowledge from a large teacher to a much smaller student over a set of data. We argue that the selection of thee three key components, namely teacher, training data, and learning objective, is cr…
Knowledge distillation (KD) is a highly promising method for mitigating the computational problems of pre-trained language models (PLMs). Among various KD approaches, Intermediate Layer Distillation (ILD) has been a de facto standard KD method with its performance efficacy in the NLP field. In this paper, we find that existing ILD methods are prone to overfitting to training datasets, although these methods transfer more information than the original KD. Next, we present the simple observations to mitigate the overfitting of ILD: distilling only the last Transformer layer and conducting ILD on…
University of California, San Diego（加州大学圣地亚哥分校）；Beihang University（北京航空航天大学）；Microsoft Research Asia（微软亚洲研究院）
In this paper, we propose a novel model compression approach to effectively compress BERT by progressive module replacing. Our approach first divides the original BERT into several modules and builds their compact substitutes. Then, we randomly replace the original modules with their substitutes to train the compact modules to mimic the behavior of the original modules. We progressively increase the probability of replacement through the training. In this way, our approach brings a deeper level of interaction between the original and compact models. Compared to the previous knowledge distillat…
University of Waterloo（滑铁卢大学）；Salesforce Research（赛富时研究院）
Pretrained language models have led to significant performance gains in many NLP tasks. However, the intensive computing resources to train such models remain an issue. Knowledge distillation alleviates this problem by learning a light-weight student model. So far the distillation approaches are all task-specific. In this paper, we explore knowledge distillation under the multi-task learning setting. The student is jointly distilled across different tasks. It acquires more general representation capacity through multi-tasking distillation and can be further fine-tuned to improve the model in t…
Microsoft Cognitive Services Research（微软认知服务研究院）；University of California, Berkeley（加州大学伯克利分校）；University of Edinburgh（爱丁堡大学）
In this paper we apply self-knowledge distillation to text summarization which we argue can alleviate problems with maximum-likelihood training on single reference and noisy datasets. Instead of relying on one-hot annotation labels, our student summarization model is trained with guidance from a teacher which generates smoothed labels to help regularize training. Furthermore, to better model uncertainty during training, we introduce multiple noise signals for both teacher and student models. We demonstrate experimentally on three benchmarks that our framework boosts the performance of both pre…
University of Massachusetts Amherst（马萨诸塞大学阿默斯特分校）；Amazon Alexa（亚马逊 Alexa）
The current state-of-the-art task-oriented semantic parsing models use BERT or RoBERTa as pretrained encoders; these models have huge memory footprints. This poses a challenge to their deployment for voice assistants such as Amazon Alexa and Google Assistant on edge devices with limited memory budgets. We propose to learn compositional code embeddings to greatly reduce the sizes of BERT-base and RoBERTa-base. We also apply the technique to DistilBERT, ALBERT-base, and ALBERT-large, three already compressed BERT variants which attain similar state-of-the-art performances on semantic parsing wit…
Task-agnostic knowledge distillation attempts to address the problem of deploying large pretrained language model in resource-constrained scenarios by compressing a large pretrained model called teacher into a smaller one called student such that the student can be directly finetuned on downstream tasks and retains comparable performance. However, we empirically find that there is a generalization gap between the student and the teacher in existing methods. In this work, we show that we can leverage multi-task learning in task-agnostic distillation to advance the generalization of the resulted…
Modular Transformers: Compressing Transformers into Modularized Layers for Flexible Efficient Inference
ETH Zurich（苏黎世联邦理工学院）；Allen Institute for AI（艾伦人工智能研究所）；Paul G. Allen School of Computer Science & Engineering, University of Washington（华盛顿大学保罗·G·艾伦计算机科学与工程学院）
Max Planck Institute for Psycholinguistics（马克斯·普朗克心理语言学研究所）；University of Ulm（乌尔姆大学）
Large pretrained language models (PreLMs) are revolutionizing natural language processing across all benchmarks. However, their sheer size is prohibitive for small laboratories or for deployment on mobile devices. Approaches like pruning and distillation reduce the model size but typically retain the same model architecture. In contrast, we explore distilling PreLMs into a different, more efficient architecture, Continual Multiplication of Words (CMOW), which embeds each word as a matrix and uses matrix multiplication to encode sequences. We extend the CMOW architecture and its CMOW/CBOW-Hybri…
Tsinghua University（清华大学）；Microsoft Research Asia（微软亚洲研究院）
Pre-trained language models (PLMs) achieve great success in NLP. However, their huge model sizes hinder their applications in many practical systems. Knowledge distillation is a popular technique to compress PLMs, which learns a small student model from a large teacher PLM. However, the knowledge learned from a single teacher may be limited and even biased, resulting in low-quality student model. In this paper, we propose a multi-teacher knowledge distillation framework named MT-BERT for pre-trained language model compression, which can train high-quality student model from multiple teacher PL…
Sentence-Level or Token-Level? A Comprehensive Study on Knowledge Distillation
Shenyang Institute of Computing Technology, Chinese Academy of Sciences（中国科学院沈阳计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；University of Science and Technology of China（中国科学技术大学）
Huawei Noah’s Ark Lab（华为诺亚方舟实验室）；University of British Columbia（不列颠哥伦比亚大学）
Knowledge Distillation (KD) is a common knowledge transfer algorithm used for model compression across a variety of deep learning based natural language processing (NLP) solutions. In its regular manifestations, KD requires access to the teacher's training data for knowledge transfer to the student network. However, privacy concerns, data regulations and proprietary reasons may prevent access to such data. We present, to the best of our knowledge, the first work on Zero-Shot Knowledge Distillation for NLP, where the student learns from the much larger teacher without any task specific data. Ou…
Knowledge distillation~(KD) has been proved effective for compressing large-scale pre-trained language models. However, existing methods conduct KD statically, e.g., the student model aligns its output distribution to that of a selected teacher model on the pre-defined training dataset. In this paper, we explore whether a dynamic knowledge distillation that empowers the student to adjust the learning procedure according to its competency, regarding the student performance and learning efficiency. We explore the dynamical adjustments on three aspects: teacher model adoption, data selection, and…
Large pre-trained transformer-based language models have achieved impressive results on a wide range of NLP tasks. In the past few years, Knowledge Distillation(KD) has become a popular paradigm to compress a computationally expensive model to a resource-efficient lightweight model. However, most KD algorithms, especially in NLP, rely on the accessibility of the original training dataset, which may be unavailable due to privacy issues. To tackle this problem, we propose a novel two-stage data-free distillation method, named Adversarial self-Supervised Data-Free Distillation (AS-DFD), which is …
Columbia University（哥伦比亚大学）；Microsoft Research（微软研究院）；University of California, Los Angeles（加利福尼亚大学洛杉矶分校）
Cross-modal encoders for vision-language (VL) tasks are often pretrained with carefully curated vision-language datasets. While these datasets reach an order of 10 million samples, the labor cost is prohibitive to scale further. Conversely, unimodal encoders are pretrained with simpler annotations that are less cost-prohibitive, achieving scales of hundreds of millions to billions. As a result, unimodal encoders have achieved state-of-art (SOTA) on many downstream tasks. However, challenges remain when applying to VL tasks. The pretraining data is not optimal for cross-modal architectures and …
University of California, Santa Barbara（加州大学圣巴巴拉分校）；Ping An Life Insurance Company of China, Ltd.（中国平安人寿保险股份有限公司）
Recently developed large pre-trained language models, e.g., BERT, have achieved remarkable performance in many downstream natural language processing applications. These pre-trained language models often contain hundreds of millions of parameters and suffer from high computation and latency in real-world applications. It is desirable to reduce the computation overhead of the models for fast training and inference while keeping the model performance in downstream applications. Several lines of work utilize knowledge distillation to compress the teacher model to a smaller student model. However,…
Feature Alignment and Representation Transfer in Knowledge Distillation for Large Language Models
Pingtan Research Institute of Xiamen University（厦门大学滨海研究院）；Imperial College London（伦敦帝国理工学院）；University of Sussex（苏塞克斯大学）；Purdue University（普渡大学）；University of Liverpool（利物浦大学）；JTB Technology Corp.（JTB技术公司）；Emory University（埃默里大学）；The University of Texas at Dallas（德克萨斯大学达拉斯分校）；Kyoto University（京都大学）；AppCubic；Georgia Institute of Technology（佐治亚理工学院）；AI Agent Lab（AI代理实验室）
UC Berkeley（加州大学伯克利分校）；Shanghai Jiao Tong University（上海交通大学）；University of Chicago（芝加哥大学）
Although pretrained Transformers such as BERT achieve high accuracy on in-distribution examples, do they generalize to new distributions? We systematically measure out-of-distribution (OOD) generalization for seven NLP datasets by constructing a new robustness benchmark with realistic distribution shifts. We measure the generalization of previous models including bag-of-words models, ConvNets, and LSTMs, and we show that pretrained Transformers' performance declines are substantially smaller. Pretrained transformers are also more effective at detecting anomalous or OOD examples, while many pre…
Topic models are often used to identify human-interpretable topics to help make sense of large document collections. We use knowledge distillation to combine the best attributes of probabilistic topic models and pretrained transformers. Our modular method can be straightforwardly applied with any neural topic model to improve topic quality, which we demonstrate using two models having disparate architectures, obtaining state-of-the-art topic coherence. We show that our adaptable framework not only improves performance in the aggregate over all estimated topics, as is commonly reported, but als…
This paper studies compressing pre-trained language models, like BERT (Devlin et al.,2019), via teacher-student knowledge distillation. Previous works usually force the student model to strictly mimic the smoothed labels predicted by the teacher BERT. As an alternative, we propose a new method for BERT distillation, i.e., asking the teacher to generate smoothed word ids, rather than labels, for teaching the student model in knowledge distillation. We call this kind of methodTextSmoothing. Practically, we use the softmax prediction of the Masked Language Model(MLM) in BERT to generate word dist…
Teacher-student knowledge distillation is a popular technique for compressing today's prevailing large language models into manageable sizes that fit low-latency downstream applications. Both the teacher and the choice of transfer set used for distillation are crucial ingredients in creating a high quality student. Yet, the generic corpora used to pretrain the teacher and the corpora associated with the downstream target domain are often significantly different, which raises a natural question: should the student be distilled over the generic corpora, so as to learn from high-quality teacher p…
Is Smaller Always Faster? Tradeoffs in Compressing Self-Supervised Speech Transformers
Graduate Institute of Communication Engineering, National Taiwan University, Taiwan（台湾国立台湾大学通信工程研究所）；University of California, Los Angeles, United States（美国加州大学洛杉矶分校）；University of Washington, United States（美国华盛顿大学）；University of Edinburgh, United Kingdom（英国爱丁堡大学）
Since deep learning became a key player in natural language processing (NLP), many deep learning models have been showing remarkable performances in a variety of NLP tasks, and in some cases, they are even outperforming humans. Such high performance can be explained by efficient knowledge representation of deep learning models. While many methods have been proposed to learn more efficient representation, knowledge distillation from pretrained deep networks suggest that we can use more information from the soft target probability to train other neural networks. In this paper, we propose a new k…
