Code for injecting knowledge for longformer training

[NtaylorOX]

Hi,

Great work!

I wonder if there is the code for the actual training/knowledge injection using umls with the contrastive loss function available?

Would be useful to see how that works in order to replicate the approach for other datasets

Thanks

[mgh1]

I second this request!