Prompt Learning via Meta-Regularization

Benchmark Model Rank Results
few-shot-domain-generalization-on-imagenet-sketchProMetaR#3Accuracy: 49.55
prompt-engineering-on-food-101ProMetaR#4Harmonic mean: 91.34
prompt-engineering-on-ucf101ProMetaR#6Harmonic mean: 83.25
prompt-engineering-on-oxford-102-flowerProMetaR#8Harmonic mean: 86.70
prompt-engineering-on-stanford-carsProMetaR#10Harmonic mean: 76.72
prompt-engineering-on-sun397ProMetaR#11Harmonic mean: 80.82
prompt-engineering-on-fgvc-aircraftProMetaR#12Harmonic mean: 40.25
prompt-engineering-on-imagenetProMetaR#12Harmonic mean: 74.09
prompt-engineering-on-dtdProMetaR#13Harmonic mean: 72.31
prompt-engineering-on-eurosatProMetaR#14Harmonic mean: 85.30
prompt-engineering-on-oxford-iiit-pet-datasetProMetaR#17Harmonic mean: 96.49
prompt-engineering-on-caltech-101ProMetaR#22Harmonic mean: 96.16