[first post, eeeek!]
Hi folks!
I’m new to the party here but wanted to see if any of you have tips or thoughts on how to save models from the SparkXGBClassifier. I tried the code in the docs on a model after fitting it and it wrote two subdirectories ‘model’ and ‘metadata’ which both contained only empty _SUCCESS files. Following the docs to then read the model from this path returned ‘ValueError: RDD is empty’.
I filed an issue about it since I saw one other person on stack overflow with the same problem six months ago, but am wondering if there are maybe some settings I need to tinker with to get it working?
Anyway, party on.