Guolin Ke
Guolin Ke
Avoiding Exception "Check failed: (best_split_info.right_count) > (0) at ..." with a regression task
is that possible to provide an exproduce example, by a sub-feature (or even with subrow), so that we can debug with.
I think we should benchmark for the training speed regards to this change.
I think compile option is much easier. Another solution is using template, and auto choose data_size_t according to #data .
@mjaysonnn GPU version cannot support categorical features with high cardinality. You can fix it by split one categorical feature into multi categorical features.
@clinchergt yeah, it is the number of unique categorical values.
I think the old GPU/CUDA version will be abandoned. also cc @shiyu1994 to follow up on this issue.
Hi, @jameslamb is the CI job normal now?
we would like to call the voting here, to prioritize these requests. If you think a feature request is very necessary for you, you can vote for it by the...
@peixin-lin can you provide a reproduce example?