guolinke Follow

guolinke Follow

🍉

Being busy, sorry for the late response.

Guolin Ke guolinke

🍉

Being busy, sorry for the late response.

Creator & Maintainer of LightGBM.

Follow

480 followers · 41 following · 87

Highlights

Arctic Code Vault Contributor

Organizations

guolinke/README.md

Pinned

microsoft/LightGBM

A fast, distributed, high performance gradient boosting (GBT, GBDT, GBRT, GBM or MART) framework based on decision tree algorithms, used for ranking, classification and many other machine learning …

C++ 11.6k 3k
motefly/DeepGBM

Implementation for the paper "DeepGBM: A Deep Learning Framework Distilled by GBDT for Online Prediction Tasks", which has been accepted by KDD'2019.

Python 492 104
TUPE

Transformer with Untied Positional Encoding (TUPE). Code of paper "Rethinking Positional Encoding in Language Pre-training". Improve existing models like BERT.

Python 70 8

607 contributions in the last year

Activity overview

Contributed to microsoft/LightGBM, guolinke/TUPE, guolinke/pytorch-docker and 5 other repositories

Contribution activity

October 1, 2020

guolinke has no activity yet for this period.

September 2020

guolinke/tokenizers Rust Sep 26
guolinke/guolinke Sep 11
guolinke/category_encoders Python Sep 10
guolinke/recommenders Python Sep 1

Created a pull request in microsoft/LightGBM that received 12 comments

fix address alignment, required by cran

+173 −111 • 12 comments

fix warnings Sep 15
stable multi-threading sum reduction Sep 12
improve subfeature_bynode Sep 12
disable monotone constraint in objective functions with renew_tree_output Sep 11
[docs] Simplify the python installation instruction Sep 11
bump version Sep 1
v3.0.0 release Sep 1

add on-the-fly frequency encoder Sep 11
support k fold in TargetEncoder Sep 10

[python] Use ctypes for parameters of DLL functions for Dataset Sep 30
Use ctypes to pass parameters (fixes #3398) Sep 30
fix address alignment, required by cran Sep 30
Fix SWIG methods that return char** Sep 29
fix warnings Sep 27
[R-package] Fix incorrect num_iterations, early_stopping_round in saved models (fixes #2208, #2468) Sep 27
[R-package] use keyword arguments in internal function calls Sep 24
Improve performance of path smoothing Sep 20
Auc mu weights Sep 20
Average precision score Sep 20
fix sparse multiclass local feature contributions and add test Sep 20
stable multi-threading sum reduction Sep 20
Pr4 advanced method monotone constraints Sep 15
Fix typo in ResetConfig Sep 13
disable monotone constraint in objective functions with renew_tree_output Sep 12
improve subfeature_bynode Sep 12
[R-package] remove unnecessary comments Sep 11
Avoid segment fault in ResetConfig for GBDT in prediction (fix #3317) Sep 9
CTR and Count values for categorical features Sep 9
Add support for CUDA-based GPU build Sep 9
Trees with linear models at leaves Sep 3
Build integrated Python package library Sep 3
[R-package] added support for first_metric_only (fixes #2368) Sep 2
[R-package] bump version Sep 1
bump version Sep 1

support k fold in TargetEncoder Sep 10

Created an issue in huggingface/tokenizers that received 11 comments

"the" token is splitted to "t" "h" "e" in large scale corpus

I train the bpe (BERTWordPiece) by my own, using the RoBERTa 160GB data. However, I found all "the" is broken. I check the learned vocab.txt, and f…

11 comments

bug when using special token with uppercase Sep 26

[GPU] Result of LightGBM GPU seems strange Sep 12

[python] suppress the warning about categorical feature override Sep 11

shorten-method cannot work Sep 8

78 contributions in private repositories Sep 1 – Sep 30

You can’t perform that action at this time.