srowen commented on a change in pull request #27546: [SPARK-30773][ML]Support
NativeBlas for level-1 routines
URL: https://github.com/apache/spark/pull/27546#discussion_r386032301
##########
File path: core/src/main/scala/org/apache/spark/internal/config/package.scala
##########
@@ -1541,4 +1541,10 @@ package object config {
.bytesConf(ByteUnit.BYTE)
.createOptional
+ private[spark] val NATIVE_L1_THRESHOLD =
+ ConfigBuilder("spark.ml.blas.nativeL1Threshold")
+ .doc("vector size for logic change between f2jBLAS and natveBLAS in " +
Review comment:
"Minimum vector size to use native BLAS libraries (if available) for Level 1
routines rather than Java implementation"
I'd also actually remove the "ml" from the property name as it isn't in an
ML package, and isn't specific to ML. However we can document it in the context
of ML as that is its only real use right now.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
