Search results

[python-package] Add decision_function() to LGBMClassifier

feature request

Status: Open.

#7158 In Microsoft/LightGBM;

· CoderWota opened on Feb 21, 2026

[python-package] Can a LGBMClassifier be refit?

question

Status: Open.

#7157 In Microsoft/LightGBM;

· bhvieira opened on Feb 20, 2026

[python-package] Refit doesnt work if the booster only has one tree

bug

Status: Open.

#7156 In Microsoft/LightGBM;

· sidhant007 opened on Feb 18, 2026

[python-package] Booster.get_split_value_histogram() is not implemented in the expected way

question

Status: Open.

#7155 In Microsoft/LightGBM;

· cafferychen opened on Feb 18, 2026

Let lightgbm opencl module support "quantization gradient" function

feature request

gpu (OpenCL)Issue is related to the OpenCL-based GPU variant.Issue is related to the OpenCL-based GPU variant.

Status: Open.

#7154 In Microsoft/LightGBM;

· shengyanli1982 opened on Feb 18, 2026

Misleading classification performance when falling back to weighted priors with certain class distributions

question

Status: Open.

#7153 In Microsoft/LightGBM;

· JNewmanUEA opened on Feb 10, 2026

[CUDA] Tweedie metric receives raw log-space scores without exp() conversion, breaking early stopping

bug

effectiveness

gpu (CUDA)Issue is related to the CUDA GPU variant.Issue is related to the CUDA GPU variant.

Status: Open.

#7152 In Microsoft/LightGBM;

· Whisdangist opened on Feb 7, 2026

WeightedPercentileFun uses wrong segment and can return values outside [min(y), max(y)]

effectiveness

question

Status: Open.

#7151 In Microsoft/LightGBM;

· hanataba0217 opened on Feb 6, 2026

[python-package] Vendor <code>libgomp.so</code> in Linux wheels for manylinux compliance?

feature request

question

Status: Open.

#7141 In Microsoft/LightGBM;

· rdbisme opened on Jan 30, 2026

[CUDA] SIGFPE (Floating point exception) with discrete data when n_unique_values * n_features exceeds threshold

bug

gpu (CUDA)Issue is related to the CUDA GPU variant.Issue is related to the CUDA GPU variant.

Status: Open.

#7122 In Microsoft/LightGBM;

· hiromuhana opened on Jan 9, 2026

[CUDA] Segfault when using int64 sparse matrix indices with <code>predict(pred_contrib=True)</code>

bug

gpu (CUDA)Issue is related to the CUDA GPU variant.Issue is related to the CUDA GPU variant.

Status: Open.

#7101 In Microsoft/LightGBM;

· nicklamiller opened on Dec 9, 2025

[R-package] MSVC compile failure (R, Windows)

bug

r-package

Status: Open.

#7098 In Microsoft/LightGBM;

· GabeAl opened on Dec 5, 2025