What is the size of parameters in FFM

[KennyTC]

In FFM, I see the following parameters.

self.embedding_size = params['embedding_size']
self.field_size = params['field_size']
self.feature_size = params['feature_size']

feature_idx = features["feature_idx"]
feature_values = features["feature_values"]      

If I have 17 columns, each column is numeric range from 0 to 99. Now I want to embed each column to k latent vectors (e.g. k = 4)
I guess in my case
self.field_size = 17,
self. embedding_size = 4
self.feature_size = 100

What is features["feature_idx"] and features["feature_values"]? and what is their shapes?