Visual Question Answering On Amber

Accuracy

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	Accuracy	F1
rlaif-v-aligning-mllms-through-open-source-ai	88	90.9