v0.12.0
NVIDIA/garakv0.12.0Jul 1, 2025by jmartin-tech
AI Summary
Adds support for audio NIM models and probes, introduces result summary objects for reporting, and adds progress indicators for translation tasks. The octo generator was removed.
Key Highlights
- Audio NIM model and audio probes support
- Result summary object for reporting
- Progress indicators for translation tasks
- Google Translate language provider
New Features
- Audio NIM model and probes
- Result summary object
- Google Translate provider
- Progress indicators
Full Release Notes
## What's Changed ### New plugins * Add audio NIM model and audio probes by @erickgalinkin in https://github.com/NVIDIA/garak/pull/1163 * Leakreplay refactor by @dchiitmalla in https://github.com/NVIDIA/garak/pull/1264 * probes: refactor fact snippet mixin by @leondz in https://github.com/NVIDIA/garak/pull/1187 ### New features * reporting: result summary object by @leondz in https://github.com/NVIDIA/garak/pull/1245 * rm octo generator by @leondz in https://github.com/NVIDIA/garak/pull/1248 * UX: add progress indicators for translation tasks by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1257 * Feature: add google translate language provider by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1232 * update LocalDataPath and testing for python 3.13 support by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1188 ### Documentation * docs: correct reference to langproviders configuration by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1253 * chore: The bare .active has too much blast radius by @mikemckiernan in https://github.com/NVIDIA/garak/pull/1262 * docs: Address RST issues by @mikemckiernan in https://github.com/NVIDIA/garak/pull/1263 * chore: Update pre-commit hooks and use RST comment by @mikemckiernan in https://github.com/NVIDIA/garak/pull/1267 * docs: mistral syntax fix and minor code cleanup by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1270 * chore: Reformat docstrings to reduce docs issues by @mikemckiernan in https://github.com/NVIDIA/garak/pull/1268 ### Tuning & fixes * data correlation during calibration missed rename by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1240 * bug fix: enable extended detectors, module-specified encoding payloads by default by @leondz in https://github.com/NVIDIA/garak/pull/1238 * adjust calling convention for perf_stats.py by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1246 * remove modality redundant check by @jmartin-tech in https://github.com/NVIDIA/garak/pull/1251 * Llava tests by @dchiitmalla in https://github.com/NVIDIA/garak/pull/1256 * add action for closing stale PRs/issues by @leondz in https://github.com/NVIDIA/garak/pull/1269 * update refusal prompt by @katherine-luna in https://github.com/NVIDIA/garak/pull/1083 * add stale exempt tags by @leondz in https://github.com/NVIDIA/garak/pull/1272 * actions: add 'high priortiy' to never-stale label list by @leondz in https://github.com/NVIDIA/garak/pull/1274 * reset tiers for `Complete` `leakreplay` probes by @leondz in https://github.com/NVIDIA/garak/pull/1278 ## New Contributors * @mikemckiernan made their first contribution in https://github.com/NVIDIA/garak/pull/1262 * @dchiitmalla made their first contribution in https://github.com/NVIDIA/garak/pull/1256 * @katherine-luna made their first contribution in https://github.com/NVIDIA/garak/pull/1083 **Full Changelog**: https://github.com/NVIDIA/garak/compare/v0.11.0...v0.12.0