One Embedder, Any Task: Instruction-Finetuned Text Embeddings
This provides a more flexible and efficient solution for NLP practitioners needing task-specific embeddings, though it is incremental as it builds on prior instruction-based methods.
The authors tackled the problem of creating a single text embedder adaptable to diverse tasks without retraining by introducing INSTRUCTOR, which embeds text with task instructions and achieved state-of-the-art performance with a 3.4% average improvement on 70 evaluation datasets.
We introduce INSTRUCTOR, a new method for computing text embeddings given task instructions: every text input is embedded together with instructions explaining the use case (e.g., task and domain descriptions). Unlike encoders from prior work that are more specialized, INSTRUCTOR is a single embedder that can generate text embeddings tailored to different downstream tasks and domains, without any further training. We first annotate instructions for 330 diverse tasks and train INSTRUCTOR on this multitask mixture with a contrastive loss. We evaluate INSTRUCTOR on 70 embedding evaluation tasks (66 of which are unseen during training), ranging from classification and information retrieval to semantic textual similarity and text generation evaluation. INSTRUCTOR, while having an order of magnitude fewer parameters than the previous best model, achieves state-of-the-art performance, with an average improvement of 3.4% compared to the previous best results on the 70 diverse datasets. Our analysis suggests that INSTRUCTOR is robust to changes in instructions, and that instruction finetuning mitigates the challenge of training a single model on diverse datasets. Our model, code, and data are available at https://instructor-embedding.github.io.