feat: LLM accuracy testing framework with 17-test catalog

Framework runs a category/subcategory test catalog against multiple LLMs through an OpenAI-compatible endpoint. Deterministic evaluators score each response. Reports render as table, markdown, or JSON.
2026-10-15 01:07:26 +00:00 · 2026-08-17 01:29:45 +01:00
commit 1d51d4cbbd
68 changed files with 7089 additions and 0 deletions
@@ -0,0 +1,10 @@
+module github.com/lukaszraczylo/llm-testbench
+
+go 1.25.0
+
+toolchain go1.26.6
+
+require (
+	golang.org/x/sync v0.22.0
+	gopkg.in/yaml.v3 v3.0.1
+)