Alan Akbik
Alan Akbik
Thanks @lixiu911 even a very small amount of data would be enough if we can reproduce the error with it. Or different public data that shows the error. We just...
Hello @i4meta yes our new version of the SequenceTagger will now always infer the full BIO/BIOES tag set for any tag in the training data. Regarding the data: best via...
@fmafelipe what Flair version are you on?
We are on Flair 0.11.3 now, perhaps you could test with this version? If the error persists, could you supply a minimal working example with data to reproduce this error?
Thanks for raising this issue @josebarross! @whoisjones can you check if this works?
Hello @dobbersc @helpmefindaname very interesting results and discussion! Regarding renaming labels: With the new corpus logic, renaming of any label is now easy and requires only to define a `label_name_map`....
Thanks for sharing these interesting results @helpmefindaname! I wonder if the masking of the "remainder" NER tags could be a problem since they share many subtokens with HEAD and TAIL...
Awesome @EmilStenstrom thanks for fixing this so quickly! And thanks @helpmefindaname for testing!
Hello @AliHaiderAhmad001 thanks for using Flair! The `TransformerWordEmbeddings` class has default handling for words split into multiple subwords which you control with the `subtoken_pooling` parameter (your choices are "first", "last",...
@boegel it was intentional since they are only minor updates, but we should probably at least always tag the code for each release.