Pg_agents on Yonk-Labshttps://yonk.dev/tags/pg_agents/Recent content in Pg_agents on Yonk-LabsHugo -- gohugo.ioen© 2026 Yonk-LabsMon, 13 Apr 2026 00:00:00 +0000Training Sets for LoRA: How to Teach a 4B Model to Write Postgres SQL Without Cryinghttps://yonk.dev/blog/training-sets-for-lora-nl2sql/Mon, 13 Apr 2026 00:00:00 +0000https://yonk.dev/blog/training-sets-for-lora-nl2sql/A layered training corpus — domain pairs, public ballast, and a reusable Postgres syntax corpus — is 80% of the work for a NL2SQL LoRA. The training config is YAML and patience.Your LLM Doesn't Know What 'Revenue' Means. Here's How to Fix That.https://yonk.dev/blog/building-your-semantic-layer/Sun, 05 Apr 2026 00:00:00 +0000https://yonk.dev/blog/building-your-semantic-layer/A step-by-step walkthrough of building a PostgreSQL semantic layer in pg_agents — crawl the schema, enrich it, define your vocabulary, lock down categoricals, and promote the queries that work.Why LLMs Alone Fail at NL2SQL (And What Actually Fixes It)https://yonk.dev/blog/why-llms-fail-at-nl2sql/Fri, 03 Apr 2026 00:00:00 +0000https://yonk.dev/blog/why-llms-fail-at-nl2sql/Raw LLMs hit 10-20% accuracy on real enterprise schemas with cryptic column names and tribal-knowledge joins. Here’s why, and the semantic-layer fix that takes you from toy to production.