lmms-eval
4,052One-for-all multimodal evaluation toolkit across text, image, video, and audio tasks.
DetectionPost-generationFull HITLExperimentalTier 2BEnvironment
Maturity tier details
Computed tier: 2B(raw score 4.44). Effective tier: 2B.
Facets:documented_in=0evaluated_on=0integration_surfaces=1paper_citations=0recent_commit=2stars=1.44
Source of truth:pipeline/maturity/score.py
Description
LMMs-Eval is the EvolvingLMMs-Lab's multimodal evaluation harness — a multimodal counterpart to lm-evaluation-harness. Implements 100+ benchmarks across text, image, video, and audio tasks in a consistent interface.
Why it's included
Fills the multimodal-eval cell. Differentiated from lm-evaluation-harness (text-only) and lighteval by its native multimodal scope.
Addresses failure modes
- Fabrication — Generating reference to non-existent APIs, packages, or symbols.
- Logic error — Correct API use, incorrect logic.
Techniques
multimodal-benchmarksunified-harness