facebookresearch/unibench
Python Library to evaluate VLM models' robustness across diverse benchmarks
Jupyter NotebookNOASSERTION
Issues
- 0
Error evaluating the sun397 dataset
#10 opened by mseitzer - 1
No local evaluation is happening
#9 opened by chaitanyakrishna1248 - 3
Dataset is not found.
#7 opened by PrettyMagnolia - 3
Error in unibench list_benchmarks: ValueError when converting 'zero-shot' to float
#6 opened by jun297 - 1
- 1
Adversarial attacks
#3 opened by HashmatShadab