From 614a1856e842d49c9fe91bfa1752dfe211ce7445 Mon Sep 17 00:00:00 2001 From: =?utf8?q?J=C3=A9r=C3=B4me=20Benoit?= Date: Mon, 27 Jan 2025 15:36:56 +0100 Subject: [PATCH] perf: fine tune LightGBM optuna params MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit Signed-off-by: Jérôme Benoit --- .../freqaimodels/LightGBMRegressorQuickAdapterV35.py | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/quickadapter/user_data/freqaimodels/LightGBMRegressorQuickAdapterV35.py b/quickadapter/user_data/freqaimodels/LightGBMRegressorQuickAdapterV35.py index 5fe43c6..ac4ecfe 100644 --- a/quickadapter/user_data/freqaimodels/LightGBMRegressorQuickAdapterV35.py +++ b/quickadapter/user_data/freqaimodels/LightGBMRegressorQuickAdapterV35.py @@ -69,7 +69,10 @@ class LightGBMRegressorQuickAdapterV35(BaseRegressionModel): start = time.time() if optuna_hyperopt: pruner = optuna.pruners.MedianPruner(n_warmup_steps=5) - study = optuna.create_study(pruner=pruner, direction="minimize") + study = optuna.create_study( + pruner=pruner, + direction="minimize", + ) study.optimize( lambda trial: objective( trial, @@ -222,9 +225,9 @@ def objective(trial, X, y, weights, X_test, y_test, params): study_params = { "objective": "rmse", "n_estimators": trial.suggest_int("n_estimators", 100, 800), - "num_leaves": trial.suggest_int("num_leaves", 20, 3000, step=10), + "num_leaves": trial.suggest_int("num_leaves", 2, 256), "learning_rate": trial.suggest_float("learning_rate", 1e-3, 0.3, log=True), - "min_child_samples": trial.suggest_int("min_child_samples", 10, 200), + "min_child_samples": trial.suggest_int("min_child_samples", 5, 100), "subsample": trial.suggest_float("subsample", 0.6, 1.0), "colsample_bytree": trial.suggest_float("colsample_bytree", 0.6, 1.0), "reg_alpha": trial.suggest_float("reg_alpha", 1e-8, 10.0, log=True), -- 2.43.0