Toolformer: Language Models Can Teach Themselves to Use Tools
This addresses the limitation of large language models in handling specific functionalities, enabling them to leverage external tools without sacrificing core abilities, which is incremental but impactful for enhancing model utility.
The paper tackles the problem of language models struggling with basic tasks like arithmetic and factual lookup by introducing Toolformer, a model that teaches itself to use external tools via APIs, achieving substantially improved zero-shot performance across various tasks, often competitive with much larger models.
Language models (LMs) exhibit remarkable abilities to solve new tasks from just a few examples or textual instructions, especially at scale. They also, paradoxically, struggle with basic functionality, such as arithmetic or factual lookup, where much simpler and smaller models excel. In this paper, we show that LMs can teach themselves to use external tools via simple APIs and achieve the best of both worlds. We introduce Toolformer, a model trained to decide which APIs to call, when to call them, what arguments to pass, and how to best incorporate the results into future token prediction. This is done in a self-supervised way, requiring nothing more than a handful of demonstrations for each API. We incorporate a range of tools, including a calculator, a Q\&A system, two different search engines, a translation system, and a calendar. Toolformer achieves substantially improved zero-shot performance across a variety of downstream tasks, often competitive with much larger models, without sacrificing its core language modeling abilities.