Github user avulanov commented on the pull request:
https://github.com/apache/spark/pull/1290#issuecomment-54757556
@elyast I am playing with various settings right now, use higher learning
rate or more iterations
I also thought that breeze will be faster but according to our tests
gradient computation in breeze is in par with bgreeven implementation, and
there is additional overhead of rolling/unrolling matrices to/from weight
vector. Moreover, using native BLAS/LAPACK through breeze doesn't speed
anything up. Interestingly, this corresponds to graphs in
https://github.com/fommil/netlib-java. I will be happy if you prove the
opposite. That said, code with breeze is easier to read.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
