Thailand

Pre-trained language models like BERT are performant in a wide range of natural language tasks. However, they are resource exhaustive and computationally expensive for industrial scenarios. Thus, early exits are adopted at each layer of BERT to perform adaptive computation by predicting easier samples with the first few layers to speed up the inference. In this work, to improve efficiency without performance drop, we propose a novel training scheme called Learned Early Exit for BERT (LeeBERT). First, we ask each exit to learn from each other, rather than learning only from the last layer. Second, the weights of different loss terms are learned, thus balancing off different objectives. We formulate the optimization of LeeBERT as a bi-level optimization problem, and we propose a novel cross-level optimization (CLO) algorithm to improve the optimization results. Experiments on the GLUE benchmark show that our proposed methods improve the performance of the state-of-the-art (SOTA) early exit methods for pre-trained models.

IJCNLP-AACL 2021

LeeBERT: Learned Early Exit for BERT with cross-level optimization

early exiting

knowledge distillation

bert

**Welcome to ACL-IJCNLP 2021!**

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP). 

As in previous years, the program of the conference includes a poster session, tutorials, workshops and demonstrations in addition to the main conference.


We were able to keep the registration fees similar to those charged for the virtual ACL 2020. The one fee allows attendance at the main conference and any/all tutorials and workshops. These fees would be $125 Regular Early and $175 Regular Late; $50 Student Early and $75 Student Late. Early registration closes at midnight July 11, 2021 (Eastern Daylight Time).

**Reminder:** It is ACL’s policy that at least one author of each accepted paper (including ACL Finding papers) must register for the conference.

**Reminder2:** Underline site will open closer to the event. If you already registered you will receive access detail

Registration is now open

The great event is jointly organized by the Association for Computational Linguistics (ACL) and Asian Federation of Natural Language Processing (AFNLP).

technical paper

Most studies on word-level Quality Estimation (QE) of machine translation focus on language-specific models. The obvious disadvantages of these approaches are the need for labelled data for each language pair and the high cost required to maintain several language-specific models. To overcome these problems, we explore different approaches to multilingual, word-level QE. We show that multilingual QE models perform on par with the current language-specific models. In the cases of zero-shot and few-shot QE, we demonstrate that it is possible to accurately predict word-level quality for any given new language pair from models trained on other language pairs. Our findings suggest that the word-level QE models based on powerful pre-trained transformers that we propose in this paper generalise well across languages, making them more useful in real-world scenarios.

An Exploratory Analysis of Multilingual Word-Level Quality Estimation with Cross-Lingual Transformers

Large-scale models for learning fixed-dimensional cross-lingual sentence representations like Large-scale models for learning fixed-dimensional cross-lingual sentence representations like LASER (Artetxe and Schwenk, 2019b) lead to significant improvement in performance on downstream tasks. However, further increases and modifications based on such large-scale models are usually impractical due to memory limitations. In this work, we introduce a lightweight dual-transformer architecture with just 2 layers for generating memory-efficient cross-lingual sentence representations. We explore different training tasks and observe that current cross-lingual training tasks leave a lot to be desired for this shallow architecture. To ameliorate this, we propose a novel cross-lingual language model, which combines the existing single-word masked language model with the newly proposed cross-lingual token-level reconstruction task. We further augment the training task by the introduction of two computationally-lite sentence-level contrastive learning tasks to enhance the alignment of cross-lingual sentence representation space, which compensates for the learning bottleneck of the lightweight transformer for generative tasks. Our comparisons with competing models on cross-lingual sentence retrieval and multilingual document classification confirm the effectiveness of the newly proposed training tasks for a shallow model.

Lightweight Cross-Lingual Sentence Representation Learning

Transformers are not suited for processing long documents, due to their quadratically increasing memory and time consumption. Simply truncating a long document or applying the sparse attention mechanism will incur the context fragmentation problem or lead to an inferior modeling capability against comparable model sizes. In this paper, we propose ERNIE-Doc, a document-level language pretraining model based on Recurrence Transformers. Two well-designed techniques, namely the retrospective feed mechanism and the enhanced recurrence mechanism, enable ERNIE-Doc, which has a much longer effective context length, to capture the contextual information of a complete document. We pretrain ERNIE-Doc to explicitly learn the relationships among segments with an additional document-aware segment-reordering objective. Various experiments were conducted on both English and Chinese document-level tasks. ERNIE-Doc improved the state-of-the-art language modeling result of perplexity to 16.8 on WikiText-103. Moreover, it outperformed competitive pretraining models by a large margin on most language understanding tasks, such as text classification and question answering.

ERNIE-Doc: A Retrospective Long-Document Modeling Transformer

Recently, knowledge distillation (KD) has shown great success in BERT compression. Instead of only learning from the teacher's soft label as in conventional KD, researchers find that the rich information contained in the hidden layers of BERT is conducive to the student's performance. To better exploit the hidden knowledge, a common practice is to force the student to deeply mimic the teacher's hidden states of all the tokens in a layer-wise manner. In this paper, however, we observe that although distilling the teacher's hidden state knowledge (HSK) is helpful, the performance gain (marginal utility) diminishes quickly as more HSK is distilled. To understand this effect, we conduct a series of analysis. Specifically, we divide the HSK of BERT into three dimensions, namely depth, length and width. We first investigate a variety of strategies to extract crucial knowledge for each single dimension and then jointly compress the three dimensions. In this way, we show that 1) the student's performance can be improved by extracting and distilling the crucial HSK, and 2) using a tiny fraction of HSK can achieve the same performance as extensive HSK distillation. Based on the second finding, we further propose an efficient KD paradigm to compress BERT, which does not require loading the teacher during the training of student. For two kinds of student models and computing devices, the proposed KD paradigm gives rise to training speedup of 2.7x ~ 3.4x.

Marginal Utility Diminishes: Exploring the Minimum Knowledge for BERT Knowledge Distillation

Lifelong learning (LL) aims to train a neural network on a stream of tasks while retaining knowledge from previous tasks. However, many prior attempts in NLP still suffer from the catastrophic forgetting issue, where the model completely forgets what it just learned in the previous tasks. In this paper, we introduce Rational LAMOL, a novel end-to-end LL framework for language models. In order to alleviate catastrophic forgetting, Rational LAMOL enhances LAMOL, a recent LL model, by applying critical freezing guided by human rationales. When the human rationales are not available, we propose exploiting unsupervised generated rationales as substitutions. In the experiment, we tested Rational LAMOL on permutations of three datasets from the ERASER benchmark. The results show that our proposed framework outperformed vanilla LAMOL on most permutations. Furthermore, unsupervised rationale generation was able to consistently improve the overall LL performance from the baseline without relying on human-annotated rationales.

Rational LAMOL: A Rationale-based Lifelong Learning Framework

Natural language processing (NLP) often faces the problem of data diversity such as different domains, themes, styles, and so on. Therefore, a single language model (LM) is insufficient to learn all knowledge from diverse samples. To solve this problem, we firstly propose an autoencoding topic model with a mixture prior (mATM) to perform clustering for the data, where the clusters defined in semantic space describes the data diversity. Having obtained the clustering assignment for each sample, we develop the ensemble LM (EnsLM) with the technique of weight modulation. Specifically, EnsLM contains a backbone that is adjusted by a few modulated weights to fit for different sample clusters. 
As a result, the backbone learns the shared knowledge among all clusters while modulated weights extract the cluster-specific features. EnsLM can be trained jointly with mATM with a flexible LM backbone. We evaluate the effectiveness of both mATM and EnsLM on various tasks.

EnsLM: Ensemble Language Model for Data Diversity by Semantic Clustering

Open Information Extraction (OpenIE) aims to extract structured relational tuples (subject, relation, object) from sentences, and plays a critical role in many NLP applications. Existing solutions perform extraction at sentence level, without referring to any additional contextual information. In reality, however, a sentence typically exists as part of a document rather than standalone; we often need to access relevant contextual information around the sentence before we can accurately interpret it.
As there is no document-level context-aware OpenIE dataset available, we manually annotate 800 sentences from 80 documents in two domains (Healthcare and Transportation) to form a DocOIE dataset for evaluation.
In addition, we propose DocIE, a document-level context-aware OpenIE model. Our experimental results demonstrate that incorporating document-level context is helpful in improving OpenIE performance. Both the DocOIE dataset and DocIE model are available online.

DocOIE: A Document-level Context-Aware Dataset for OpenIE

Medical entity retrieval is an integral component for understanding and communicating information across various health systems. Current approaches tend to work well on specific medical domains but generalize poorly to unseen sub-specialties. This is of increasing concern under a public health crisis as new medical conditions and drug treatments come to light frequently. Zero-shot retrieval is challenging due to the high degree of ambiguity and variability in medical corpora, making it difficult to build an accurate similarity measure between mentions and concepts. Medical knowledge graphs (KG), however, contain rich semantics including large numbers of synonyms as well as its curated graphical structures. To take advantage of this valuable information, we propose a suite of learning tasks designed for training efficient zero-shot entity retrieval models. Without requiring any human annotation, our knowledge graph enriched architecture significantly outperforms common zero-shot benchmarks including BM25 and Clinical BERT with 7% to 30% higher recall across multiple major medical ontologies, such as UMLS, SNOMED, and ICD-10.

Zero-shot Medical Entity Retrieval without Annotation: Learning From Rich Knowledge Graph Semantics

Traditional toxicity detection models have focused on the single utterance level without deeper understanding of context. We introduce CONDA, a new dataset for in-game toxic language detection enabling joint intent classification and slot filling analysis, which is the core task of Natural Language Understanding (NLU). The dataset consists of 45K utterances from 12K conversations from the chat logs of 1.9K completed Dota 2 matches. We propose a robust dual semantic-level toxicity framework, which handles utterance and token-level patterns, and rich contextual chatting history. Accompanying the dataset is a thorough in-game toxicity analysis, which provides comprehensive understanding of context at utterance, token, and dual levels. Inspired by NLU, we also apply its metrics to the toxicity detection tasks for assessing toxicity and game-specific aspects. We evaluate strong NLU models on CONDA, providing fine-grained results for different intent classes and slot classes. Furthermore, we examine the coverage of toxicity nature in our dataset by comparing it with other toxicity datasets.

CONDA: a CONtextual Dual-Annotated dataset for in-game toxicity understanding and detection

Event detection (ED) aims at detecting event trigger words in sentences and classifying them into specific event types. In real-world applications, ED typically does not have sufficient labeled data, thus can be formulated as a few-shot learning problem. To tackle the issue of low sample diversity in few-shot ED, we propose a novel knowledge-based few-shot event detection method which uses a definition-based encoder to introduce external event knowledge as the knowledge prior of event types. Furthermore, as external knowledge typically provides limited and imperfect coverage of event types, we introduce an adaptive knowledge-enhanced Bayesian meta-learning method to dynamically adjust the knowledge prior of event types. Experiments show our method consistently and substantially outperforms a number of baselines by at least 15 absolute $F_1$ points under the same few-shot settings.

LeeBERT: Learned Early Exit for BERT with cross-level optimization

Downloads

Next from IJCNLP-AACL 2021

An Exploratory Analysis of Multilingual Word-Level Quality Estimation with Cross-Lingual Transformers

Similar lecture

BERTifying the Hidden Markov Model for Multi-Source Weakly Supervised Named Entity Recognition

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

LeeBERT: Learned Early Exit for BERT with cross-level optimization

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from IJCNLP-AACL 2021

An Exploratory Analysis of Multilingual Word-Level Quality Estimation with Cross-Lingual Transformers

Similar lecture

BERTifying the Hidden Markov Model for Multi-Source Weakly Supervised Named Entity Recognition

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads