Closed tylerjthomas9 closed 1 year ago
To be honest I don't really understand the intended use cases for save
vs serialize
.
I am not sure the best way to compare if loading in a booster worked correctly other than comparing predictions.
Yup, that seems like the safest way to me.
Only comment is that for CI/CD, git and reproducibility sake model_file
should be created with mktemp
or something like that, otherwise looks good. :+1:
:eyes:
Can this please be tagged to have the fix for #126 ?
I will try to figure out tagging later today now that I have permissions. Tagbot is also broken for some reason so I'd like to get that fixed. I don't know exactly when I'll get it tagged but it's the very next thing I intend to do on this package.
I've found that TagBot has just stopped working on a couple of repos I manage for no discernible reason. Updating the relevant rsa keys seemed to fix the issue.
I don't think I can do that, the only thing I have access to is the config yaml. I don't have access to the settings page.
In the meantime I have tagged in the registry if anyone knows specifically anything I can do to fix tagbot, please let me know.
Added some testing to verify that the save/load/serialize functionality is working as expected. I am not sure the best way to compare if loading in a booster worked correctly other than comparing predictions.
This PR includes changes from https://github.com/dmlc/XGBoost.jl/pull/125, and the typo mentioned here https://github.com/dmlc/XGBoost.jl/issues/126