Deep Multimodal Neural Architecture Search

Benchmark Model Rank Results
visual-question-answering-on-vqa-v2-test-stdSingle, w/o VLP#19overall: 73.86yes/no: 89.46number: 58.62other: 63.78