A programming framework for agentic AI
train at least one iter when not trained (#336)
Chi Wang434586e2e2d7
diff --git a/flaml/model.py b/flaml/model.pyindex fe427b284..95f2c8f68 100644--- a/flaml/model.py+++ b/flaml/model.py@@ -868,26 +868,24 @@ class LGBMEstimator(BaseEstimator):)if trained and max_iter <= self.params[self.ITER_HP]:return time.time() - start_time- self.params[self.ITER_HP] = max_iter- if self.params[self.ITER_HP] > 0:- if self.HAS_CALLBACK:- self._fit(- X_train,- y_train,- callbacks=self._callbacks(start_time, deadline),- **kwargs,- )- best_iteration = (- self._model.get_booster().best_iteration- if isinstance(self, XGBoostSklearnEstimator)- else self._model.best_iteration_- )- if best_iteration is not None:- self._model.set_params(n_estimators=best_iteration + 1)- else:- self._fit(X_train, y_train, **kwargs)+ # when not trained, train at least one iter+ self.params[self.ITER_HP] = max(max_iter, 1)+ if self.HAS_CALLBACK:+ self._fit(+ X_train,+ y_train,+ callbacks=self._callbacks(start_time, deadline),+ **kwargs,+ )+ best_iteration = (+ self._model.get_booster().best_iteration+ if isinstance(self, XGBoostSklearnEstimator)+ else self._model.best_iteration_+ )+ if best_iteration is not None:+ self._model.set_params(n_estimators=best_iteration + 1)else:- self.params[self.ITER_HP] = self._model.n_estimators+ self._fit(X_train, y_train, **kwargs)train_time = time.time() - start_timereturn train_timediff --git a/flaml/version.py b/flaml/version.pyindex 3e2f46a3a..d69d16e98 100644--- a/flaml/version.py+++ b/flaml/version.py@@ -1 +1 @@-__version__ = "0.9.0"+__version__ = "0.9.1"