Gecko: Versatile Text Embeddings Distilled from Large Language Models
This work addresses the need for efficient and high-performance text embeddings for retrieval tasks, offering a significant improvement over existing models in terms of compactness and versatility.
The paper tackles the problem of creating compact text embedding models by distilling knowledge from large language models, achieving state-of-the-art retrieval performance on the Massive Text Embedding Benchmark with Gecko's 256-dimensional embeddings outperforming existing 768-dimensional ones and its 768-dimensional version scoring 66.31, competing with much larger models.
We present Gecko, a compact and versatile text embedding model. Gecko achieves strong retrieval performance by leveraging a key idea: distilling knowledge from large language models (LLMs) into a retriever. Our two-step distillation process begins with generating diverse, synthetic paired data using an LLM. Next, we further refine the data quality by retrieving a set of candidate passages for each query, and relabeling the positive and hard negative passages using the same LLM. The effectiveness of our approach is demonstrated by the compactness of the Gecko. On the Massive Text Embedding Benchmark (MTEB), Gecko with 256 embedding dimensions outperforms all existing entries with 768 embedding size. Gecko with 768 embedding dimensions achieves an average score of 66.31, competing with 7x larger models and 5x higher dimensional embeddings.