A programming framework for agentic AI
fixing use_ray in automl.py (#531)
Xueqing Liuc1e1299855dc
diff --git a/flaml/automl.py b/flaml/automl.pyindex 28f1c36df..90d9ebf94 100644--- a/flaml/automl.py+++ b/flaml/automl.py@@ -1891,7 +1891,7 @@ class AutoML(BaseEstimator):for estimator in self.estimator_list:search_state = self._search_states[estimator]if not hasattr(search_state, "training_function"):- if self._use_ray:+ if self._use_ray is not False:from ray.tune import with_parameterssearch_state.training_function = with_parameters(@@ -2292,7 +2292,7 @@ class AutoML(BaseEstimator):self._use_ray = use_ray or n_concurrent_trials > 1# use the following condition if we have an estimation of average_trial_time and average_trial_overhead# self._use_ray = use_ray or n_concurrent_trials > ( average_trail_time + average_trial_overhead) / (average_trial_time)- if self._use_ray:+ if self._use_ray is not False:import rayn_cpus = use_ray and ray.available_resources()["CPU"] or os.cpu_count()@@ -2358,7 +2358,7 @@ class AutoML(BaseEstimator):self._state.retrain_final = (retrain_full is Trueand eval_method == "holdout"- and (self._state.X_val is None or self._use_ray)+ and (self._state.X_val is None or self._use_ray is not False)or eval_method == "cv"and (max_iter > 0 or retrain_full is True)or max_iter == 1@@ -2557,7 +2557,9 @@ class AutoML(BaseEstimator):if hpo_method != "auto"else ("bs"- if n_concurrent_trials > 1 or self._use_ray and len(estimator_list) > 1+ if n_concurrent_trials > 1+ or self._use_ray is not False+ and len(estimator_list) > 1else "cfo"))@@ -3114,7 +3116,7 @@ class AutoML(BaseEstimator):if isinstance(state.init_config, dict)else state.init_config[0])- elif not self._use_ray:+ elif self._use_ray is False:self._search_sequential()else:self._search_parallel()