Provides LLM-as-judge and code-based evaluators for scoring LLM outputs, with built-in templates for hallucination, relevance, and toxicity detection.

Required Ruby Version

>= 3.4

Authors

Matthew Iverson

Versions

  1. 0.3.13

    1. source

      1. 0.3.13 source March 30, 2026 (21.5 KB)
  2. 0.3.11

    1. source

      1. 0.3.11 source March 30, 2026 (21.5 KB)
  3. 0.3.10

    1. source

      1. 0.3.10 source March 29, 2026 (21 KB)
  4. 0.3.9

    1. source

      1. 0.3.9 source March 29, 2026 (21 KB)
  5. 0.3.8

    1. source

      1. 0.3.8 source March 27, 2026 (20.5 KB)
Show all versions (13 total)

Pushed by

LegionIO

SHA 256 checksum

← Previous version