{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/strata-self-training-with-task-augmentation","title":"STraTA: Self-Training with Task Augmentation for Better Few-shot Learning","arxiv_id":"2109.06270","date":"2021-09-13","proceeding":"EMNLP 2021 11","authors":["Tu Vu","Minh-Thang Luong","Quoc V. Le","Grady Simon","Mohit Iyyer"],"abstract":"Despite their recent successes in tackling many NLP tasks, large-scale pre-trained language models do not perform as well in few-shot settings where only a handful of training examples are available. To address this shortcoming, we propose STraTA, which stands for Self-Training with Task Augmentation, an approach that builds on two key ideas for effective leverage of unlabeled data. First, STraTA uses task augmentation, a novel technique that synthesizes a large amount of data for auxiliary-task fine-tuning from target-task unlabeled texts. Second, STraTA performs self-training by further fine-tuning the strong base model created by task augmentation on a broad distribution of pseudo-labeled data. Our experiments demonstrate that STraTA can substantially improve sample efficiency across 12 few-shot benchmarks. Remarkably, on the SST-2 sentiment dataset, STraTA, with only 8 training examples per class, achieves comparable results to standard fine-tuning with 67K training examples. Our analyses reveal that task augmentation and self-training are both complementary and independently effective.","url_abs":"https://arxiv.org/abs/2109.06270v2","url_pdf":"https://arxiv.org/pdf/2109.06270v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"strata-self-training-with-task-augmentation","repo_url":"https://github.com/google-research/google-research/tree/master/STraTA","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"jax","reach":null}],"tasks":[{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"few-shot-nli","task_name":"Few-Shot NLI"},{"task_slug":null,"task_name":"SST-2"}],"methods":[{"method_slug":"strata","method_name":"STraTA"}],"datasets_introduced":[],"methods_introduced":[{"slug":"strata","name":"STraTA","full_name":"Self-Training with Task Augmentation"}],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2109.06270","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}