M3DBench

M3DBench introduces a comprehensive 3D instruction-following dataset with support for interleaved multi-modal prompts. Furthermore, M3DBench provides a new benchmark to assess large models across 3D vision-centric tasks.

GitHub

54 stars
5 watching
2 forks
last commit: 10 months ago
3dbenchmarkdatasetinstruction-tuninglanguage-modelllmmlmmulti-modalpromptvlm