AncaSarb

@guzzijones @Ben-Epstein We have two custom transformers now available in pyspark based on the work from @juhoautio and @lucagiovagnoli, the code changes are here https://github.com/combust/mleap/pull/666, perhaps you can take a...

Hey @femibyte, we have these converters for e.g. https://github.com/combust/mleap/blob/e177e5816250ff0dfae7f82e84317f288a107f31/mleap-spark-base/src/main/scala/ml/combust/mleap/spark/SparkSupport.scala#L63 that would do this for you. This is currently in Scala, but I'll update the issue once I've gotten it to...

At the moment, we have scikit-learn pinned to < 0.20, we would need to upgrade the scikit-learn version to get this working. Happy to work with you on doing this...

Sure thing, I'll take a look!

Currently, mleap-runtime can only be invoked from Scala/Java/JVM based language. What you can do though, is to have either a grpc client or http client in Python, calling a scoring...

@voganrc @talalryz if i'm not mistaken, @lucagiovagnoli already made this switch in master, right? could i please ask you to double check if we can close this issue? thank you!

@gizemerd do you have an idea how big your model is? so that i can try to reproduce the issue.

See if you can follow https://github.com/combust/mleap/tree/master/mleap-spring-boot and use the most recently released version 0.15.0, the easiest would be to do this, as mentioned in that README.md file ```The previous scoring...

Could you please provide a bit more detail on your use case? In which case you've found the row by row prediction slow? Thank you!

In our use case, we had to support model building/training not just in scikit-learn, but also in Spark and Tensorflow, so MLeap helped in this case, because at scoring time,...