Are the leaderboard results from `0-shot` settings?

OpenGVLab / LAMM

[NeurIPS 2023 Datasets and Benchmarks Track] LAMM: Multi-Modal Large Language Models and Applications as AI Agents

https://openlamm.github.io/

286 stars 15 forks source link

Closed zhimin-z closed 7 months ago

zhimin-z commented 7 months ago

Coach257 commented 7 months ago

Yes, all the results are from 0-shot settings, except for ICL evaluation.