The ALCHEmist: Automated Labeling 500x CHEaper Than LLM Data Annotators
This addresses the problem of expensive and inflexible data annotation for machine learning practitioners, offering a cost-effective alternative to LLM-based annotation.
The paper tackles the high cost and static nature of using large pretrained models as data annotators by proposing a system that generates programs to produce labels instead, achieving comparable or better performance with a 12.9% average improvement and reducing labeling costs by approximately 500x.
Large pretrained models can be used as annotators, helping replace or augment crowdworkers and enabling distilling generalist models into smaller specialist models. Unfortunately, this comes at a cost: employing top-of-the-line models often requires paying thousands of dollars for API calls, while the resulting datasets are static and challenging to audit. To address these challenges, we propose a simple alternative: rather than directly querying labels from pretrained models, we task models to generate programs that can produce labels. These programs can be stored and applied locally, re-used and extended, and cost orders of magnitude less. Our system, Alchemist, obtains comparable to or better performance than large language model-based annotation in a range of tasks for a fraction of the cost: on average, improvements amount to a 12.9% enhancement while the total labeling costs across all datasets are reduced by a factor of approximately 500x.