Github user avulanov commented on the pull request:
https://github.com/apache/spark/pull/1290#issuecomment-101091531
I did another small test to compare new implementation
(https://github.com/avulanov/spark/tree/ann-interface-gemm) performance with
the current one in this branch.
* Cluster configuration (OpenBLAS is used)
* 6 machines (Xeon 3.3GHz 4 cores, 16GB RAM) with 5 workers total,
* mnist8m dataset, persist in memory
* Network topology 784x10 (no hidden layer = logistic regression)
* LBFGS optimizer, 40 steps (epochs), tolerance 1e-4, batch size = 100
* Average time on 3 runs
* Accuracy on mnist test set: 0.9076
Implementation | time, hh:mm:ss | Accuracy |
--------|------------|------ |-----
Current ANN | 00:03:09 | 0.9076
New ANN | 00:02:33 | 0.9076
LogisticRegressionWithLBFGS* | 00:04:11 | 0.9087
*Does not take advantage of batch computations and native BLAS
---
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]
