[Solved] LigthGBM error

[continuing-vincent]

Hi,

I tried to submit the “basic_submission” public notebook with a lightgbm model instead of an xgboost.

Everything works fine locally, even the crunch.test(…). However, it raises the following error on cloud :
lightgbm.basic.LightGBMError: scikit-learn is required for lightgbm.sklearn. You must install scikit-learn and restart your session to use this module.

SOLUTION FOUND : just import sklearn in the notebook. It’s not very practical to manually import dependancies though.

[prime-paulius]

LightGBM Error - within train function (one of the 2 required functions to submit) when calling lightgbm.train gettint the following error (few weeks ago that was not the issue):

[enzo]

@continuing-vincent

SOLUTION FOUND : just import sklearn in the notebook. It’s not very practical to manually import dependancies though.

Good to know that you found the issue.
Its a limitation of the Notebook importer sorry, we can’t really guess what dependencies you are gonna use, so we have to find them from your imports.

[enzo]

@prime-paulius

LightGBM Error - within train function (one of the 2 required functions to submit) when calling lightgbm.train gettint the following error (few weeks ago that was not the issue):

Create a new post and please share your Run ID if you need help.