Embedding Model Selection for a Production Search System

text-embedding-ada-002 is not always the right choice. Benchmarking five open-weight models on our domain-specific retrieval task produced a surprising ranking.

Overview

This note is part of the field-notes archive generated for this site. The summary below is the published excerpt; you can expand the full write-up anytime in the CMS.

Series

Part of ML in Production (installment 5).

Related notes

• Rag Vs Finetuning Decision Framework
• Ml Drift Monitoring Production

Tags

• embeddings
• search
• retrieval
• machine-learning
• nlp