The original code for the paper "How to train your MAML" along with a replication of the original "Model Agnostic Meta Learning" (MAML) paper in Pytorch.
Hi,
Thanks for your great work MAML++.
I'm having a problem with multi-GPU. When I input
bash experiment_scripts.sh 0 1 2
to run the training, it seems that only the first GPU ID is being passed, while the others are being ignored.
I tried specifying the GPU IDs directly in the script.sh, but it didn't work. Is this a bug, or am I missing something?
Thanks!
Due to how pytorch has changed over time, multi GPU is no longer supported on this codebase. If I find time in the future (or if GPT5 is good enough to do this for me) I'll try to bring it back.
Hi, Thanks for your great work MAML++. I'm having a problem with multi-GPU. When I input
to run the training, it seems that only the first GPU ID is being passed, while the others are being ignored. I tried specifying the GPU IDs directly in the script.sh, but it didn't work. Is this a bug, or am I missing something? Thanks!