Laurae

Results 36 comments of Laurae

@ogrisel Note that LightGBM number of threads scale with the number of columns. Higgs dataset does not have enough columns for 48 threads (it will underestimate the scalability which gives...

@spedygiorgio LightGBM R installation changed recently therefore this is currently obsolete until I setup a new way of installing from `lgbdl`. See here: https://github.com/Microsoft/LightGBM/pull/1499

@spedygiorgio Please test the new version bb36351 of [lgbdl](https://github.com/Laurae2/lgbdl) (current master).

bb36351 should take the first letter from the temporary directory now.

10M dataset is too small for that many cores, it gets quickly diminishing returns: https://public.tableau.com/views/GradientBoostedTreesMegaBenchmark2/TabularTime?:showVizHome=no ![image](https://user-images.githubusercontent.com/9083669/27685641-68bd49bc-5ccf-11e7-9c46-9e814419e605.png) ![image](https://user-images.githubusercontent.com/9083669/27685738-dccd397a-5ccf-11e7-8700-f96da3c6e0d6.png)

@guolinke Do you know how to use LightGBM MPI locally in R? @szilard yeah, 10M is a lot. Most cases are around 1K to 1M.