You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The new NodeNorm provides better information on identifiers in MedMentions, so retraining BioMegatron with these new identifiers and Biolink types might help BioMegatron make better type determinations.
The text was updated successfully, but these errors were encountered:
Chris suggested to me today that distinguishing between genes and proteins might be confusing for BioMegatron, since they are often used in similar contexts. At some point, we should try training BioMegatron without distinguishing between genes and proteins and see if that improves its accuracy (assuming it's needed after training it on the new NodeNorm).
The new NodeNorm provides better information on identifiers in MedMentions, so retraining BioMegatron with these new identifiers and Biolink types might help BioMegatron make better type determinations.
The text was updated successfully, but these errors were encountered: