MaPLe: Multi-modal Prompt Learning

Benchmark Model Rank Results
prompt-engineering-on-caltech-101MaPLe#11Harmonic mean: 96.02
prompt-engineering-on-dtdMaPLe#12Harmonic mean: 68.16
prompt-engineering-on-eurosatMaPLe#10Harmonic mean: 82.35
prompt-engineering-on-fgvc-aircraftMaPLe#11Harmonic mean: 36.50
prompt-engineering-on-food-101MaPLe#3Harmonic mean: 91.38
prompt-engineering-on-imagenetMaPLe#13Harmonic mean: 73.47
prompt-engineering-on-imagenet-aMaPLe#4Top-1 accuracy %: 50.90
prompt-engineering-on-imagenet-rMaPLe#7Top-1 accuracy %: 76.98
prompt-engineering-on-imagenet-sMaPLe#7Top-1 accuracy %: 49.15
prompt-engineering-on-imagenet-v2MaPLe#6Top-1 accuracy %: 64.07
prompt-engineering-on-oxford-102-flowerMaPLe#12Harmonic mean: 82.56
prompt-engineering-on-oxford-iiit-pet-datasetMaPLe#6Harmonic mean: 96.58
prompt-engineering-on-stanford-cars-1MaPLe#12Harmonic mean: 73.47
prompt-engineering-on-sun397MaPLe#11Harmonic mean: 79.75
prompt-engineering-on-ucf101MaPLe#11Harmonic mean: 80.82