It is just an implementation of Model Agnostic Meta learning and First order approximation of it using learn2learn library in Pytorch on Omniglot dataset.
Use the following command
pip install torch torchvision torchaudio matplotlib learn2learn scipy
To open the ipynb file you would need jupyter notebook
pip install jupyter notebook
If you have a GPU please use the code as it is
If you don't have a GPU installed
Change the following flag in the ipynb file
cuda=False
This code also times the execution time using time library and the jupyter code as been run for 10000 steps, if you are intrested in seeing the difference and learning more about the MAML amd FOMAML.
El Psy Kongroo my friend
If any issues open a Github query @CaffineAddic