Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on Feb 17, 2026

• feat(model): export Verbosity and LinearSelectionMode types

  Show description for b78923d
  farhan-syah

  committed
  b78923d

  Copy full SHA for b78923d

• test: update tests for validated configuration methods

  Show description for c6b56aa
  farhan-syah

  committed
  c6b56aa

  Copy full SHA for c6b56aa

• chore: remove unused imports and cleanup whitespace

  Show description for 81aee1f
  farhan-syah

  committed
  81aee1f

  Copy full SHA for 81aee1f

• refactor(universal): simplify prediction logic and cleanup unused code

  Show description for 3269107
  farhan-syah

  committed
  3269107

  Copy full SHA for 3269107

• feat(model): add custom loss and validation dataset enhancements

  Show description for ab0a810
  farhan-syah

  committed
  ab0a810

  Copy full SHA for ab0a810

• refactor(pipeline): improve type handling and encoding consistency

  Show description for 4fc746e
  farhan-syah

  committed
  4fc746e

  Copy full SHA for 4fc746e

• feat(universal): add learning rate and subsample validation

  Show description for 313823c
  farhan-syah

  committed
  313823c

  Copy full SHA for 313823c

• feat(model): add granular verbosity control for AutoBuilder

  Show description for 68c85d9
  farhan-syah

  committed
  68c85d9

  Copy full SHA for 68c85d9

• feat(gpu): add column subsampling support for CUDA and WebGPU backends

  Show description for 260259f
  farhan-syah

  committed
  260259f

  Copy full SHA for 260259f

• feat(dataset): add impute_value field to FeatureInfo for NaN handling

  Show description for 87d99d0
  farhan-syah

  committed
  87d99d0

  Copy full SHA for 87d99d0

Commits on Jan 15, 2026

• feat(tree): add era validation for noise pruning robustness

  Show description for b2d4910
  farhan-syah

  committed
  b2d4910

  Copy full SHA for b2d4910

• feat(tree): add post-pruning and noise pruning for improved accuracy

  Show description for 8d5c7f9
  farhan-syah

  committed
  8d5c7f9

  Copy full SHA for 8d5c7f9

• test: relax PseudoHuber comparison with tolerance

  Show description for 3e954ea
  farhan-syah

  committed
  3e954ea

  Copy full SHA for 3e954ea

• docs(preprocessing): fix import in LogitTransform example

  Show description for 45f1002
  farhan-syah

  committed
  45f1002

  Copy full SHA for 45f1002

• fix(ensemble): update split_info calls for new signature

  Show description for e6d88d0
  farhan-syah

  committed
  e6d88d0

  Copy full SHA for e6d88d0

• feat(dataset): add support for pre-split train/validation data

  Show description for e51275f
  farhan-syah

  committed
  e51275f

  Copy full SHA for e51275f

• refactor(gpu): unify split finding with SplitFinder trait

  Show description for a0590cd
  farhan-syah

  committed
  a0590cd

  Copy full SHA for a0590cd

• feat(tree): add adaptive missing value direction learning

  Show description for 5d84d3f
  farhan-syah

  committed
  5d84d3f

  Copy full SHA for 5d84d3f

Commits on Jan 14, 2026

• refactor: improve code formatting and organization

  Show description for 83629ac
  farhan-syah

  committed
  83629ac

  Copy full SHA for 83629ac

• chore(deps): upgrade rand to 0.9 and add rand_distr

  Show description for c4047f8
  farhan-syah

  committed
  c4047f8

  Copy full SHA for c4047f8

• test: add comprehensive test suite for core features

  Show description for 3b28095
  farhan-syah

  committed
  3b28095

  Copy full SHA for 3b28095

• feat(gbdt): add GbdtPreset::Robust for high-dimensional and noisy data

  Show description for a6e89ad
  farhan-syah

  committed
  a6e89ad

  Copy full SHA for a6e89ad

• feat(builder): add early validation when all features are dropped

  Show description for 2eaa664
  farhan-syah

  committed
  2eaa664

  Copy full SHA for 2eaa664

• fix(profiler): improve ID detection to avoid false positives on continuous features

  Show description for e2a759f
  farhan-syah

  committed
  e2a759f

  Copy full SHA for e2a759f

• refactor(tuner): split autotuner into focused submodules with quality improvements

  Show description for ca3001e
  farhan-syah

  committed
  ca3001e

  Copy full SHA for ca3001e

• refactor: update callers to handle Result types from split functions

  Show description for 80131f7
  farhan-syah

  committed
  80131f7

  Copy full SHA for 80131f7

• feat(model): add unified predictor abstraction with proper error handling

  Show description for ae61cc2
  farhan-syah

  committed
  ae61cc2

  Copy full SHA for ae61cc2

• refactor(loss): convert loss function constructors to Result types

  Show description for 3f19d77
  farhan-syah

  committed
  3f19d77

  Copy full SHA for 3f19d77

• refactor(dataset): replace assertions with Result types in split functions

  Show description for 574649a
  farhan-syah

  committed
  574649a

  Copy full SHA for 574649a

• feat(model): add custom feature expressions and target transformation system

  Show description for f590d0c
  farhan-syah

  committed
  f590d0c

  Copy full SHA for f590d0c

Commits on Jan 12, 2026

• refactor(model): consolidate feature engineering into unified plan-based system

  Show description for 168221b
  farhan-syah

  committed
  168221b

  Copy full SHA for 168221b

• refactor(backend): improve hybrid GPU routing and observability

  Show description for c55e4fa
  farhan-syah

  committed
  c55e4fa

  Copy full SHA for c55e4fa

Commits on Jan 11, 2026

• docs: add multi-label classification quick start guide

  Show description for e96ead4
  farhan-syah

  committed
  e96ead4

  Copy full SHA for e96ead4

• test: update test suite for unified vector tree architecture

  Show description for 45a2c63
  farhan-syah

  committed
  45a2c63

  Copy full SHA for 45a2c63

• fix(model): update incremental learning for ensemble tree API

  Show description for 3a23352
  farhan-syah

  committed
  3a23352

  Copy full SHA for 3a23352