I think this line in your code assumes that there is one token required to describe each edge feature which I don't think is a fair assumption given that the edge features are a combination of two words without a space between and not a word by themselves. Please let me know what you think.
I think this line in your code assumes that there is one token required to describe each edge feature which I don't think is a fair assumption given that the edge features are a combination of two words without a space between and not a word by themselves. Please let me know what you think.