Аз ғана мысалдар бойынша тестілер
task
ID: fewshotbenchmarks
Related Papers
STraTA: Self-Training with Task Augmentation for Better Few-shot Learning
AbstractDespite 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. …...