Learned metrics such as BLEURT have in recent years become widely employed to evaluate the quality of machine translation systems. Training such metrics requires data which can be expensive and difficult to acquire, particularly for lower-resource languages. We show how knowledge can be distilled from Large Language Models (LLMs) to improve upon such learned metrics without requiring human annotators, by creating synthetic datasets which can be mixed into existing datasets, requiring only a corpus of text in the target language. We show that the performance of a BLEURT-like model on lower resource languages can be improved in this way.
相關內容
Large multilingual models trained with self-supervision achieve state-of-the-art results in a wide range of natural language processing tasks. Self-supervised pretrained models are often fine-tuned on parallel data from one or multiple language pairs for machine translation. Multilingual fine-tuning improves performance on low-resource languages but requires modifying the entire model and can be prohibitively expensive. Training a new adapter on each language pair or training a single adapter on all language pairs without updating the pretrained model has been proposed as a parameter-efficient alternative. However, the former does not permit any sharing between languages, while the latter shares parameters for all languages and is susceptible to negative interference. In this paper, we propose training language-family adapters on top of mBART-50 to facilitate cross-lingual transfer. Our approach outperforms related baselines, yielding higher translation scores on average when translating from English to 17 different low-resource languages. We also show that language-family adapters provide an effective method to translate to languages unseen during pretraining.
Many natural language processing (NLP) tasks rely on labeled data to train machine learning models to achieve high performance. However, data annotation can be a time-consuming and expensive process, especially when the task involves a large amount of data or requires specialized domains. Recently, GPT-3.5 series models have demonstrated remarkable few-shot and zero-shot ability across various NLP tasks. In this paper, we first claim that large language models (LLMs), such as GPT-3.5, can serve as an excellent crowdsourced annotator by providing them with sufficient guidance and demonstrated examples. To make LLMs to be better annotators, we propose a two-step approach, 'explain-then-annotate'. To be more precise, we begin by creating prompts for every demonstrated example, which we subsequently utilize to prompt a LLM to provide an explanation for why the specific ground truth answer/label was chosen for that particular example. Following this, we construct the few-shot chain-of-thought prompt with the self-generated explanation and employ it to annotate the unlabeled data. We conduct experiments on three tasks, including user input and keyword relevance assessment, BoolQ and WiC. The annotation results from GPT-3.5 surpasses those from crowdsourced annotation for user input and keyword relevance assessment. Additionally, for the other two tasks, GPT-3.5 achieves results that are comparable to those obtained through crowdsourced annotation.
Crowdsourcing allows running simple human intelligence tasks on a large crowd of workers, enabling solving problems for which it is difficult to formulate an algorithm or train a machine learning model in reasonable time. One of such problems is data clustering by an under-specified criterion that is simple for humans, but difficult for machines. In this demonstration paper, we build a crowdsourced system for image clustering and release its code under a free license at //github.com/Toloka/crowdclustering. Our experiments on two different image datasets, dresses from Zalando's FEIDEGGER and shoes from the Toloka Shoes Dataset, confirm that one can yield meaningful clusters with no machine learning algorithms purely with crowdsourcing.
We present new benchmarks on evaluation code generation models: MBXP and Multilingual HumanEval, and MathQA-X. These datasets cover over 10 programming languages and are generated using a scalable conversion framework that transpiles prompts and test cases from the original Python datasets into the corresponding data in the target language. Using these benchmarks, we are able to assess the performance of code generation models in a multi-lingual fashion, and discovered generalization ability of language models on out-of-domain languages, advantages of multi-lingual models over mono-lingual, the ability of few-shot prompting to teach the model new languages, and zero-shot translation abilities even on mono-lingual settings. Furthermore, we use our code generation model to perform large-scale bootstrapping to obtain synthetic canonical solutions in several languages, which can be used for other code-related evaluations such as code insertion, robustness, or summarization tasks. Overall, our benchmarks represents a significant step towards a deeper understanding of language models' code generation abilities. We publicly release our code and datasets at //github.com/amazon-research/mxeval.
The performance of abstractive text summarization has been greatly boosted by pre-trained language models recently. The main concern of existing abstractive summarization methods is the factual inconsistency problem of their generated summary. To alleviate the problem, many efforts have focused on developing effective factuality evaluation metrics based on natural language inference and question answering et al. However, they have limitations of high computational complexity and relying on annotated data. Most recently, large language models such as ChatGPT have shown strong ability in not only natural language understanding but also natural language inference. In this paper, we study the factual inconsistency evaluation ability of ChatGPT under the zero-shot setting by evaluating it on the coarse-grained and fine-grained factuality evaluation tasks including binary natural language inference (NLI), summary ranking, and consistency rating. Experimental results show that ChatGPT outperforms previous SOTA evaluation metrics on 6/9 datasets across three tasks, demonstrating its great potential for assessing factual inconsistency in the zero-shot setting. The results also highlight the importance of prompt design and the need for future efforts to address ChatGPT's limitations on evaluation bias, wrong reasoning, and hallucination.
Learning on big data brings success for artificial intelligence (AI), but the annotation and training costs are expensive. In future, learning on small data is one of the ultimate purposes of AI, which requires machines to recognize objectives and scenarios relying on small data as humans. A series of machine learning models is going on this way such as active learning, few-shot learning, deep clustering. However, there are few theoretical guarantees for their generalization performance. Moreover, most of their settings are passive, that is, the label distribution is explicitly controlled by one specified sampling scenario. This survey follows the agnostic active sampling under a PAC (Probably Approximately Correct) framework to analyze the generalization error and label complexity of learning on small data using a supervised and unsupervised fashion. With these theoretical analyses, we categorize the small data learning models from two geometric perspectives: the Euclidean and non-Euclidean (hyperbolic) mean representation, where their optimization solutions are also presented and discussed. Later, some potential learning scenarios that may benefit from small data learning are then summarized, and their potential learning scenarios are also analyzed. Finally, some challenging applications such as computer vision, natural language processing that may benefit from learning on small data are also surveyed.
Since the 1950s, machine translation (MT) has become one of the important tasks of AI and development, and has experienced several different periods and stages of development, including rule-based methods, statistical methods, and recently proposed neural network-based learning methods. Accompanying these staged leaps is the evaluation research and development of MT, especially the important role of evaluation methods in statistical translation and neural translation research. The evaluation task of MT is not only to evaluate the quality of machine translation, but also to give timely feedback to machine translation researchers on the problems existing in machine translation itself, how to improve and how to optimise. In some practical application fields, such as in the absence of reference translations, the quality estimation of machine translation plays an important role as an indicator to reveal the credibility of automatically translated target languages. This report mainly includes the following contents: a brief history of machine translation evaluation (MTE), the classification of research methods on MTE, and the the cutting-edge progress, including human evaluation, automatic evaluation, and evaluation of evaluation methods (meta-evaluation). Manual evaluation and automatic evaluation include reference-translation based and reference-translation independent participation; automatic evaluation methods include traditional n-gram string matching, models applying syntax and semantics, and deep learning models; evaluation of evaluation methods includes estimating the credibility of human evaluations, the reliability of the automatic evaluation, the reliability of the test set, etc. Advances in cutting-edge evaluation methods include task-based evaluation, using pre-trained language models based on big data, and lightweight optimisation models using distillation techniques.
This paper offers a comprehensive review of the research on Natural Language Generation (NLG) over the past two decades, especially in relation to data-to-text generation and text-to-text generation deep learning methods, as well as new applications of NLG technology. This survey aims to (a) give the latest synthesis of deep learning research on the NLG core tasks, as well as the architectures adopted in the field; (b) detail meticulously and comprehensively various NLG tasks and datasets, and draw attention to the challenges in NLG evaluation, focusing on different evaluation methods and their relationships; (c) highlight some future emphasis and relatively recent research issues that arise due to the increasing synergy between NLG and other artificial intelligence areas, such as computer vision, text and computational creativity.
The recent GPT-3 model (Brown et al., 2020) achieves remarkable few-shot performance solely by leveraging a natural-language prompt and a few task demonstrations as input context. Inspired by their findings, we study few-shot learning in a more practical scenario, where we use smaller language models for which fine-tuning is computationally efficient. We present LM-BFF--better few-shot fine-tuning of language models--a suite of simple and complementary techniques for fine-tuning language models on a small number of annotated examples. Our approach includes (1) prompt-based fine-tuning together with a novel pipeline for automating prompt generation; and (2) a refined strategy for dynamically and selectively incorporating demonstrations into each context. Finally, we present a systematic evaluation for analyzing few-shot performance on a range of NLP tasks, including classification and regression. Our experiments demonstrate that our methods combine to dramatically outperform standard fine-tuning procedures in this low resource setting, achieving up to 30% absolute improvement, and 11% on average across all tasks. Our approach makes minimal assumptions on task resources and domain expertise, and hence constitutes a strong task-agnostic method for few-shot learning.
Person Re-identification (re-id) faces two major challenges: the lack of cross-view paired training data and learning discriminative identity-sensitive and view-invariant features in the presence of large pose variations. In this work, we address both problems by proposing a novel deep person image generation model for synthesizing realistic person images conditional on pose. The model is based on a generative adversarial network (GAN) and used specifically for pose normalization in re-id, thus termed pose-normalization GAN (PN-GAN). With the synthesized images, we can learn a new type of deep re-id feature free of the influence of pose variations. We show that this feature is strong on its own and highly complementary to features learned with the original images. Importantly, we now have a model that generalizes to any new re-id dataset without the need for collecting any training data for model fine-tuning, thus making a deep re-id model truly scalable. Extensive experiments on five benchmarks show that our model outperforms the state-of-the-art models, often significantly. In particular, the features learned on Market-1501 can achieve a Rank-1 accuracy of 68.67% on VIPeR without any model fine-tuning, beating almost all existing models fine-tuned on the dataset.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191