search

katotakuma0501 / Fine-grained-NER-models

2 stars 0 forks source link

How do I use this approach for different datasets with different label levels? #2

Open asartipi13 opened 1 year ago

asartipi13 commented 1 year ago

Hi,  I hope you are doing well.  I want to apply this approach to two-level labels, and I wonder how I should choose the type_label_dim in the extract_bottom_embedding which is currently equal to 202, and the label_dim parameter for the broadcast function, which is 402 now. Besides, in the broadcast function there is a loop in the range of 2 to 202, and I wonder where 202 came from. Thanks in advance.