2.0.0

What's Changed

• Save more information on judge failures and rename final failed folder by @gabegma in #88
• Modify user simulator prompt and improve user behavioral fidelity metric by @gabegma in #89
• Add elevenlabs framework by @katstankiewicz in #90
• Improvements to the results analysis app by @fanny-riols in #77
• Fix metric bugs by @gabegma in #91
• Improve faithfulness and conversation progression for S2S by @gabegma in #93
• process elevenlabs as cascade by @katstankiewicz in #94
• Run all metrics by default by @JosephMarinier in #96
• Enhance robustness for turn-taking metrics by @nhhoang96 in #95
• Omit new names for pass@k by @gabegma in #98
• individualize override parameters per backend by @raghavm243512 in #97
• Scratch/gemini tool fix by @tara-servicenow in #99
• Bump metric version to account for recent changes by @tara-servicenow in #101
• Update nvidia websocket by @katstankiewicz in #103
• Classify Ultravox as AUDIO_LLM in get_pipeline_type by @tara-servicenow in #104
• Add assets directory in dockerfile so that background noise files are… by @tara-servicenow in #102
• Don't allow multiple domains with existing run id by @tara-servicenow in #105
• add gemini ALM support by @raghavm243512 in #106
• Update OpenAI by @katstankiewicz in #110
• Various bug fixes and improvements to metrics by @gabegma in #112
• Update documentation for ElevenLabs User Simulator and leaderboard model configs by @fanny-riols in #111
• Add metric versionning by @gabegma in #113
• Update website with v2 by @tara-servicenow in #114
• Refactor model config by @JosephMarinier in #109
• Website fixes by @tara-servicenow in #115
• Change how to run multiple domains by @JosephMarinier in #100
• Website fixes: Perturbation significance asterisks; scope domain toggle to scatter plot by @lindsaydbrin in #116
• Elevenlabs not saving by @katstankiewicz in #107
• Pr/tara/website names by @tara-servicenow in #118
• bump version for release by @katstankiewicz in #117

New Contributors

• @lindsaydbrin made their first contribution in #116

Full Changelog: 0.1.3...2.0.0