This is a deep learning library specialised for meta-learning. It uses fastai(v1) and Pytorch. Currently the meta learning DataBunch only supports from_df() functionality.
from MetaAI.models import *
from MetaAI.train import *
from MetaAI.data import *
from fastai.vision import *
from fastai.callbacks import *
import pandas as pd
data = MetaDataBunch(path='../Omniglot/Data/images_background',
df=pd.read_csv('../Omniglot/Omniglot.csv'),
label_col='class',
bs=5,
val_bs=95,
size=32,
ways=5,
shots=1)
model = resnet.resnet18()
model.fc = layers.Linear(512,5)
learn = MetaLearner.from_model(data=data,
model=model,
mode='meta_sgd',
loss_func=nn.CrossEntropyLoss(reduction='sum'),
callback_fns=[ShowGraph,
partial(ReduceLROnPlateauCallback,patience=3,factor=0.1,min_delta=5e-3)
])
learn.meta_fit(epochs=20,lr=1e-3,outer_batch_size=16) Task performed was Omniglot 5-way 1-shot. All models were trained for 20 epochs with meta lr = 1e-3. All models used 3 channel 32x32 images as input.
| Model | Meta-SGD | MAML |
|---|---|---|
| Default Net | 94.2 | 79.5 |
| Resnet18 | 71.9 | 51.9 |
| Resnet18 (pretrained) | 53.4 | 26.1 |
*MAML has a low accuracy because it needs more epochs to converge. It has a slower convergence speed than Meta-SGD.
The code was partly inspired from: https://github.com/jik0730/Meta-SGD-pytorch