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
@raphaelhoffmann Especially since the loading time of coreNLP is so (relatively) long, it seems like a better default than batch_size=1000 would be to divide lines by the number of cores (or cores*nodes for distribute). For example I just got a 2x speedup on an ec-2 node really easily here (had forgotten to set batch size first time around...). Thoughts?
The text was updated successfully, but these errors were encountered:
@raphaelhoffmann Especially since the loading time of coreNLP is so (relatively) long, it seems like a better default than
batch_size=1000
would be to divide lines by the number of cores (orcores*nodes
for distribute). For example I just got a 2x speedup on an ec-2 node really easily here (had forgotten to set batch size first time around...). Thoughts?The text was updated successfully, but these errors were encountered: