Few Shot Learning
Few-Shot Learning is a meta-learning approach that trains a model on multiple related tasks during the meta-training phase, enabling it to generalize to unseen but related tasks with only a few samples during the meta-testing phase. The method aims to learn a general representation and then train task-specific classifiers based on this representation, thereby enhancing the model's adaptability and efficiency on new tasks.
Caltech101
CaseHOLD
CR
DART
DTD
SaSPA + CAL
EuroSAT
Variational Prompt Tuning
FGVC Aircraft
Flowers-102
food101
Variational Prompt Tuning
GLUE QQP
Large COVID-19 CT scan slice dataset
MedConceptsQA
MedNLI
CoT-T5-11B (1024 Shot)
Mini-ImageNet - 1-Shot Learning
HCTransformers
Mini-ImageNet - 5-Shot Learning
Mini-Imagenet 5-way (1-shot)
HCTransformers
MR
MRPC
OxfordPets
PubMedQA
CoT-T5-11B (1024 Shot)
SST-2 Binary classification
DART
StanforCars
Stanford Cars
SUN397
UCF101
Variational Prompt Tuning