Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userThe SelectGen Challenge: Finding the Best Training Samples for Few-Shot Neural Text Generation
التحدي SelectGen: العثور على أفضل عينات تدريبية لتوليد النص العصبي القليل
460
0 0
0.0
(
0
)
Created by
Shamra Editor
Ask ChatGPT about the research
نقترح مهمة مشتركة على اختيار مثيل التدريب لعدد قليل من الجيل العصبي العصبي.أدت نماذج اللغة المحددة مسبقا على نطاق واسع إلى تحسينات مثيرة في جيل نص قليل.ومع ذلك، فإن كل العمل السابق تقريبا يطبق ما عليك سوى أخذ عينات عشوائية لتحديد مثيلات التدريب القليلة.لقد تم إيلاء الاهتمام ليس إلى استراتيجيات الاختيار وكيف سيؤثرون على أداء النموذج.دراسة استراتيجية الاختيار يمكن أن تساعدنا (1) الاستفادة القصوى من ميزانية التوضيحية الخاصة بنا في مهام المصب و (2) من أفضل النماذج الإندارية النصية ذات القليل من القصاصات.نرحب بالتقديمات التي تقدم استراتيجيات اختيارها والآثار على جودة الجيل.
We propose a shared task on training instance selection for few-shot neural text generation. Large-scale pretrained language models have led to dramatic improvements in few-shot text generation. Nonetheless, almost all previous work simply applies random sampling to select the few-shot training instances. Little to no attention has been paid to the selection strategies and how they would affect model performance. Studying the selection strategy can help us (1) make the most use of our annotation budget in downstream tasks and (2) better benchmark few-shot text generative models. We welcome submissions that present their selection strategies and the effects on the generation quality.
References used
https://aclanthology.org/
rate research
Read More
Automatic detection of critical plot information in reviews of media items poses unique challenges to both social computing and computational linguistics. In this paper we propose to cast the problem of discovering spoiler bias in online discourse as
a text simplification task. We conjecture that for an item-user pair, the simpler the user review we learn from an item summary the higher its likelihood to present a spoiler. Our neural model incorporates the advanced transformer network to rank the severity of a spoiler in user tweets. We constructed a sustainable high-quality movie dataset scraped from unsolicited review tweets and paired with a title summary and meta-data extracted from a movie specific domain. To a large extent, our quantitative and qualitative results weigh in on the performance impact of named entity presence in plot summaries. Pretrained on a split-and-rephrase corpus with knowledge distilled from English Wikipedia and fine-tuned on our movie dataset, our neural model shows to outperform both a language modeler and monolingual translation baselines.
Neural table-to-text generation models have achieved remarkable progress on an array of tasks. However, due to the data-hungry nature of neural models, their performances strongly rely on large-scale training examples, limiting their applicability in
real-world applications. To address this, we propose a new framework: Prototype-to-Generate (P2G), for table-to-text generation under the few-shot scenario. The proposed framework utilizes the retrieved prototypes, which are jointly selected by an IR system and a novel prototype selector to help the model bridging the structural gap between tables and texts. Experimental results on three benchmark datasets with three state-of-the-art models demonstrate that the proposed framework significantly improves the model performance across various evaluation metrics.
Providing pretrained language models with simple task descriptions in natural language enables them to solve some tasks in a fully unsupervised fashion. Moreover, when combined with regular learning from examples, this idea yields impressive few-shot
results for a wide range of text classification tasks. It is also a promising direction to improve data efficiency in generative settings, but there are several challenges to using a combination of task descriptions and example-based learning for text generation. In particular, it is crucial to find task descriptions that are easy to understand for the pretrained model and to ensure that it actually makes good use of them; furthermore, effective measures against overfitting have to be implemented. In this paper, we show how these challenges can be tackled: We introduce GenPET, a method for text generation that is based on pattern-exploiting training, a recent approach for combining textual instructions with supervised learning that only works for classification tasks. On several summarization and headline generation datasets, GenPET gives consistent improvements over strong baselines in few-shot settings.
Conditional text generation often requires lexical constraints, i.e., which words should or shouldn't be included in the output text. While the dominant recipe for conditional text generation has been large-scale pretrained language models that are f
inetuned on the task-specific training data, such models do not learn to follow the underlying constraints reliably, even when supervised with large amounts of task-specific examples. We propose NeuroLogic Decoding, a simple yet effective algorithm that enables neural language models -- supervised or not -- to generate fluent text while satisfying complex lexical constraints. Our approach is powerful yet efficient. It handles any set of lexical constraints that is expressible under predicate logic, while its asymptotic runtime is equivalent to conventional beam search. Empirical results on four benchmarks show that NeuroLogic Decoding outperforms previous approaches, including algorithms that handle a subset of our constraints. Moreover, we find that unsupervised models with NeuroLogic Decoding often outperform supervised models with conventional decoding, even when the latter is based on considerably larger networks. Our results suggest the limit of large-scale neural networks for fine-grained controllable generation and the promise of inference-time algorithms.
Natural Language Processing (NLP) is increasingly relying on general end-to-end systems that need to handle many different linguistic phenomena and nuances. For example, a Natural Language Inference (NLI) system has to recognize sentiment, handle num
bers, perform coreference, etc. Our solutions to complex problems are still far from perfect, so it is important to create systems that can learn to correct mistakes quickly, incrementally, and with little training data. In this work, we propose a continual few-shot learning (CFL) task, in which a system is challenged with a difficult phenomenon and asked to learn to correct mistakes with only a few (10 to 15) training examples. To this end, we first create benchmarks based on previously annotated data: two NLI (ANLI and SNLI) and one sentiment analysis (IMDB) datasets. Next, we present various baselines from diverse paradigms (e.g., memory-aware synapses and Prototypical networks) and compare them on few-shot learning and continual few-shot learning setups. Our contributions are in creating a benchmark suite and evaluation protocol for continual few-shot learning on the text classification tasks, and making several interesting observations on the behavior of similarity-based methods. We hope that our work serves as a useful starting point for future work on this important topic.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
