HyperAI

Visual Reasoning On Nlvr

Metrics

Accuracy (Dev)
Accuracy (Test-P)
Accuracy (Test-U)

Results

Performance results of various models on this benchmark

Comparison Table
Model NameAccuracy (Dev)Accuracy (Test-P)Accuracy (Test-U)
visualbert-a-simple-and-performant-baseline67.4%67%67.3%