Home

Awesome

GBML

A collection of Gradient-Based Meta-Learning Algorithms with pytorch

python3 main.py --alg=MAML
python3 main.py --alg=Reptile
python3 main.py --alg=CAVIA

Results on miniImagenet

5way 1shot5way 1shot (ours)5way 5shot5way 5shot (ours)
MAML48.70 ± 1.84%49.00 %63.11 ± 0.92%65.18 %
Reptile47.07 ± 0.26%43.40 %62.74 ± 0.37%-
CAVIA49.84 ± 0.68% (128)50.07 % (64)64.63 ± 0.53% (128)64.21 % (64)
iMAML49.30 ± 1.88%---
Meta-Curvature55.73 ± 0.94% (128)-70.30 ± 0.72% (128)-
5way 1shot5way 1shot (ours)5way 5shot5way 5shot (ours)
Meta-SGD56.58 ± 0.21%-68.84 ± 0.19%-
LEO61.76 ± 0.08%-77.59 ± 0.12%-
Meta-Curvature61.85 ± 0.10%-77.02 ± 0.11%-

Dependencies

To do