-
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
5 x decrease in performance with new Julia build #7000
Comments
Can you do |
I use the binary builds, so I guess I cannot do that? |
You could try running the fast and slow versions each with |
I can tell the slowdown is in this part: for (i, text) in enumerate(dc)
tokens::Array{UTF8String,1} = ngramize(tokenize(text), n_grams)
for token in tokens
push!(is, i)
push!(js, get!(vocabulary, token, length(vocabulary)+1))
push!(vs, 1)
end
end Will do a profile now.. |
Good:
Bad:
|
Seems likely to be |
Is the dataset available so I can try to reproduce this? |
The shootout benchmark |
Please try again. Reopen if not fixed. |
Note that if you can't reopen, you can just comment and someone will reopen it for you. |
I am currently working on a high performance text analytics package for Julia.
When today I installed the new release I got a 5 times decrease in performance with the current Julia build (
Version 0.3.0-prerelease+3225 (2014-05-27 03:45 UTC)
) when compared to an older build (Version 0.3.0-prerelease+2727 (2014-04-23 18:25 UTC)
). It is not clear why this is happening. When reinstalling the old version it works as expected. I used the windows x64 build.The code is quite long but perhaps the bottleneck is clear. Otherwise I can try to make a more simple showcase of the performance problem. I am still learning Julia so perhaps it is an obvious mistake.
The text was updated successfully, but these errors were encountered: