Closed ledlou closed 2 years ago
hi @ledlou,
did you solve the problem?
hi @ledlou,
did you solve the problem?
Yes, I did but honestly I do not even remember what was exactly the problem :) I use mlens models without any issues from that time, except for when there are remaining data with NaN values due to some coding mistake.
Sorry, I was referring to saving sub models and not having to train them again during stacking. If this was your problem, I was asking how did you solve it. There is an open issue here which includes a possible solution.
No, my original problem was simply about the saving/restoring of the final optimised model. I was quite beginner in those times :) However, the referred topic is also very useful, thank you for that!
Oh, I understand. Thank you!
How can we save fitted model and restore for subsequent tests?