Data-Efficient Finetuning Using Cross-Task Nearest Neighbors

Obtaining labeled data to train a model for a task of interest is often expensive. Prior work shows training models on multitask data augmented with task descriptions (prompts) effectively transfers knowledge to new tasks. Towards efficiently building task-specific models, we assume access to a small number (32-1000) of unlabeled target-task examples and use those to retrieve the most similar labeled examples from a large pool of multitask data augmented with prompts. Compared to the current practice of finetuning models on uniformly sampled prompted multitask data (e.g., FLAN, T0), our approach of finetuning on cross-task nearest neighbors is significantly more data-efficient. Using only 2% of the data from the P3 pool without any labeled target-task data, our models outperform strong baselines trained on all available data by 3-30% on 12 out of 14 datasets representing held-out tasks including legal and scientific document QA. Similarly, models trained on cross-task nearest neighbors from SuperNat-uralInstructions, representing about 5% of the pool, obtain comparable performance to stateof-the-art models on 12 held-out tasks from that pool. Moreover, the models produced by our approach also provide a better initialization than single multitask finetuned models for fewshot finetuning on target-task data, as shown by a 2-23% relative improvement over fewshot finetuned T0-3B models on 8 datasets. We publicly release our code. 1