Thailand

Pre-trained language models (PLMs) have achieved great success in natural language processing. Most of PLMs follow the default setting of architecture hyper-parameters (e.g., the hidden dimension is a quarter of the intermediate dimension in feed-forward sub-networks) in BERT. Few studies have been conducted to explore the design of architecture hyper-parameters in BERT, especially for the more efficient PLMs with tiny sizes, which are essential for practical deployment on resource-constrained devices. In this paper, we adopt the one-shot Neural Architecture Search (NAS) to automatically search architecture hyper-parameters. Specifically, we carefully design the techniques of one-shot learning and the search space to provide an adaptive and efficient development way of tiny PLMs for various latency constraints. We name our method AutoTinyBERT and evaluate its effectiveness on the GLUE and SQuAD benchmarks. The extensive experiments show that our method outperforms both the SOTA search-based baseline (NAS-BERT) and the SOTA distillation-based methods (such as DistilBERT, TinyBERT, MiniLM, and MobileBERT). In addition, based on the obtained architectures, we propose a more efficient development method that is even faster than the development of a single PLM. The source code and models will be publicly available upon publication.

IJCNLP-AACL 2021

AutoTinyBERT: Automatic Hyper-parameter Optimization for Efficient Pre-trained Language Models

autotinybert

efficient bert

hyper-parameter optimization

**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).

poster

Despite recent successes of large pre-trained language models in solving reasoning tasks, their inference capabilities remain opaque. We posit that such models can be made more interpretable by explicitly generating interim inference rules, and using them to guide the generation of task-specific textual outputs. In this paper we present \textsc{Coins}, a recursive inference framework that i) iteratively reads context sentences, ii) dynamically generates contextualized inference rules, encodes them, and iii) uses them to guide task-specific output generation. We apply \COINS\ to a \textit{Narrative Story Completion} task that asks a model to complete a story with missing sentences, to produce a coherent story with plausible logical connections, causal relationships, and temporal dependencies. By modularizing inference and sentence generation steps in a recurrent model, we aim to make reasoning steps and their effects on next sentence generation transparent. Our automatic and manual evaluations show that the model generates better story sentences than SOTA baselines, especially in terms of coherence. We further demonstrate improved performance over strong pre-trained LMs in generating commonsense inference rules. The recursive nature of \COINS\ holds the potential for controlled generation of longer sequences.

COINS: Dynamically Generating COntextualized Inference Rules for Narrative Story Completion

Procedural text understanding aims at tracking the states (e.g., create, move, destroy) and locations of the entities mentioned in a given paragraph. To effectively track the states and locations, it is essential to capture the rich semantic relations between entities, actions, and locations in the paragraph. Although recent works have achieved substantial progress, most of them focus on leveraging the inherent constraints or incorporating external knowledge for state prediction. The rich semantic relations in the given paragraph are largely overlooked. In this paper, we propose a novel approach (REAL) to procedural text understanding, where we build a general framework to systematically model the entity-entity, entity-action, and entity-location relations using a graph neural network. We further develop algorithms for graph construction, representation learning, and state and location tracking. We evaluate the proposed approach on two benchmark datasets, ProPara, and Recipes. The experimental results show that our method outperforms strong baselines by a large margin, i.e., 5.0% on ProPara and 3.2% on Recipes, illustrating the utility of semantic relations and the effectiveness of the graph-based reasoning model.

Reasoning over Entity-Action-Location Graph for Procedural Text Understanding

Semantic parsing is challenging due to the structure gap and the semantic gap between utterances and logical forms. In this paper, we propose an unsupervised semantic parsing method - Synchronous Semantic Decoding (SSD), which can simultaneously resolve the semantic gap and the structure gap by jointly leveraging paraphrasing and grammar-constrained decoding. Specifically, we reformulate semantic parsing as a constrained paraphrasing problem: given an utterance, our model synchronously generates its canonical utterancel and meaning representation. During synchronously decoding: the utterance paraphrasing is constrained by the structure of the logical form, therefore the canonical utterance can be paraphrased controlledly; the semantic decoding is guided by the semantics of the canonical utterance, therefore its logical form can be generated unsupervisedly. Experimental results show that SSD is a promising approach and can achieve state-of-the-art unsupervised semantic parsing performance on multiple datasets.

From Paraphrasing to Semantic Parsing: Unsupervised Semantic Parsing via Synchronous Semantic Decoding

Despite the well-developed cut-edge representation learning for language, most language representation models usually focus on specific levels of linguistic units. This work introduces universal language representation learning, i.e., embeddings of different levels of linguistic units or text with quite diverse lengths in a uniform vector space. We propose the training objective MiSAD that utilizes meaningful n-grams extracted from large unlabeled corpus by a simple but effective algorithm for pre-trained language models. Then we empirically verify that well designed pre-training scheme may effectively yield universal language representation, which will bring great convenience when handling multiple layers of linguistic objects in a unified way. Especially, our model achieves the highest accuracy on analogy tasks in different language levels and significantly improves the performance on downstream tasks in the GLUE benchmark and a question answering dataset.

Pre-training Universal Language Representation

Pre-trained language models (PrLMs) have demonstrated superior performance due to their strong ability to learn universal language representations from self-supervised pre-training. However, even with the help of the powerful PrLMs, it is still challenging to effectively capture task-related knowledge from dialogue texts which are enriched by correlations among speaker-aware utterances. In this work, we present SPIDER, Structural Pre-traIned DialoguE Reader, to capture dialogue exclusive features. To simulate the dialogue-like features, we propose two training objectives in addition to the original LM objectives: 1) utterance order restoration, which predicts the order of the permuted utterances in dialogue context; 2) sentence backbone regularization, which regularizes the model to improve the factual correctness of summarized subject-verb-object triplets. Experimental results on widely used dialogue benchmarks verify the effectiveness of the newly introduced self-supervised tasks.

Structural Pre-training for Dialogue Comprehension

We introduce data augmentation methods to utilize pretrained language models for information-seeking dialogue system. We adopt Question Answering datasets for Knowledge Identification (KI) stage, and Knowledge Grounded Dialogue datasets for Response Generation (RG) stage. Our work highlights the effectiveness of data augmentation methods and potential for practical applications.

CAiRE in DialDoc21: Data Augmentation for Information-Seeking Dialogue System

We present our entries to both subtasks of the first DialDoc shared task which focuses on the agent response prediction task in goal-oriented document-grounded dialogs. The task is split into two subtasks: predicting a span in a document that grounds an agent turn and generating an agent response based on a dialog and grounding document. In the first subtask, we restrict the set of valid spans to the ones defined in the dataset, use a biaffine classifier to model spans, and finally use an ensemble of different models. For the second subtask, we use a cascaded model which grounds the response prediction on the predicted span instead of the full document. With these approaches, we obtain significant improvements in both subtasks compared to the baseline.

Cascaded Span Extraction and Response Generation for Document-Grounded Dialog

Retrieving relevant answers from heterogeneous data formats, for given for questions, is a challenging problem. The proposed models leverage the superior predictive power of pretrained transformer models like RoBERTa, ALBERT and ELECTRA, to identify the most relevant information in an associated passage for the next agent turn. To further enhance the performance, we have explored fine-tuning models on different question answering datasets like SQuAD2.0 and Natural Questions (NQ) corpus before using ensemble techniques for combining multiple models. Our team SB_NITK ranked 6th on the leaderboard for the Knowledge Identification task, and our best ensemble model achieved an Exact score of 58.58 and an F1 score of 73.39.

Ensemble ALBERT and RoBERTa for Span Prediction in Question Answering

Most prior work on task-oriented dialogue systems are restricted to limited coverage of domain APIs. However, users oftentimes have requests that are out of the scope of these APIs. This work focuses on responding to these beyond-API-coverage user turns by incorporating external, unstructured knowledge sources. Our approach works in a pipelined manner with knowledge-seeking turn detection, knowledge selection, and response generation in sequence. We introduce novel data augmentation methods for the first two steps and demonstrate that the use of information extracted from dialogue context improves the knowledge selection and end-to-end performances. Through experiments, we achieve state-of-the-art performance for both automatic and human evaluation metrics on the DSTC9 Track 1 benchmark dataset, validating the effectiveness of our contributions.

Can I Be of Further Assistance? Using Unstructured Knowledge Access to Improve Task-oriented Conversational Modeling

Neural models for natural language generation are known to produce output which is boring, inconsistent and inappropriate. In this talk, I will argue that one way to alleviate these problems is via `grounding’. That is, encoding external knowledge which provides additional information at decoding time. I will illustrate this argument for 4 different tasks: data-to-text generation, document summarisation, visual dialogue and open-domain dialogue systems.

AutoTinyBERT: Automatic Hyper-parameter Optimization for Efficient Pre-trained Language Models

Downloads

Next from IJCNLP-AACL 2021

COINS: Dynamically Generating COntextualized Inference Rules for Narrative Story Completion

Similar lecture

Multi-Annotator Modeling to Encode Diverse Perspectives in Hate Speech Annotations