How to use Torch-meta for custom dataset?

tristandeleu / pytorch-meta

A collection of extensions and data-loaders for few-shot learning & meta-learning in PyTorch

https://tristandeleu.github.io/pytorch-meta/

MIT License

1.97k stars 256 forks source link

How to use Torch-meta for custom dataset? #137

Open karndeepsingh opened 3 years ago

karndeepsingh commented 3 years ago

Hi, I have train and test data for few classes and I want to apply few shot learning to it. How I can prepare my data for Torch-meta to train?

And Finally After model gets trained how I can get inference on new images?

Please help me. Thanks

tristandeleu commented 2 years ago

Hi, sorry for the late reply.

There is no official guide on how to use Torchmeta with your own dataset. However, I suggest you to look at this comment and the references inside; that describes the way Torchmeta works under the hood, including the 3 levels you would have to implement. My best advice is to take inspiration from the Omniglot dataset, to better understand what is going on.