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
I have tried using your parser to test with EWT English treebank, and surprisingly UAS and LAS scores are low, around 87.50 and 84.53. I have used conll2017 shared task pretrained word embeddings. Do you think this is normal or am I doing something wrong?
The text was updated successfully, but these errors were encountered:
It is normal. I also obtained similar scores on English EWT treebank. You might want to tune hyper-parameter to get better performance, rather than using default ones.
I have tried using your parser to test with EWT English treebank, and surprisingly UAS and LAS scores are low, around 87.50 and 84.53. I have used conll2017 shared task pretrained word embeddings. Do you think this is normal or am I doing something wrong?
The text was updated successfully, but these errors were encountered: