Skip to content

A programming framework for agentic AI

min eci depends on cost_attr; cost_attr in ls

Chi Wang6cb44484f6d3
diff --git a/flaml/searcher/blendsearch.py b/flaml/searcher/blendsearch.py
index ae681b62f..74682bed9 100644
--- a/flaml/searcher/blendsearch.py
+++ b/flaml/searcher/blendsearch.py
@@ -281,7 +281,7 @@ class BlendSearch(Searcher):
self._start_time = now
self._set_deadline()
if self._input_cost_attr == "auto":
- self.cost_attr = TIME_TOTAL_S
+ self.cost_attr = self._ls.cost_attr = TIME_TOTAL_S
if "metric_target" in spec:
self._metric_target = spec.get("metric_target")
if "num_samples" in spec:
@@ -820,22 +820,29 @@ class BlendSearch(Searcher):
def _select_thread(self) -> Tuple:
"""thread selector; use can_suggest to check LS availability"""
- # update priority
- now = time.time()
- min_eci = self._deadline - now
- if min_eci <= 0:
- # return -1, -1
- # keep proposing new configs assuming no budget left
- min_eci = 0
+ # calculate min_eci according to the budget left
+ min_eci = np.inf
+ if self.cost_attr == TIME_TOTAL_S:
+ now = time.time()
+ min_eci = self._deadline - now
+ if min_eci <= 0:
+ # return -1, -1
+ # keep proposing new configs assuming no budget left
+ min_eci = 0
+ elif self._num_samples and self._num_samples > 0:
+ # estimate time left according to num_samples limitation
+ num_finished = len(self._result)
+ num_proposed = num_finished + len(self._trial_proposed_by)
+ num_left = max(self._num_samples - num_proposed, 0)
+ if num_proposed > 0:
+ time_used = now - self._start_time + self._time_used
+ min_eci = min(min_eci, time_used / num_finished * num_left)
+ # print(f"{min_eci}, {time_used / num_finished * num_left}, {num_finished}, {num_left}")
elif self._num_samples and self._num_samples > 0:
- # estimate time left according to num_samples limitation
num_finished = len(self._result)
num_proposed = num_finished + len(self._trial_proposed_by)
- num_left = max(self._num_samples - num_proposed, 0)
- if num_proposed > 0:
- time_used = now - self._start_time + self._time_used
- min_eci = min(min_eci, time_used / num_finished * num_left)
- # print(f"{min_eci}, {time_used / num_finished * num_left}, {num_finished}, {num_left}")
+ min_eci = max(self._num_samples - num_proposed, 0)
+ # update priority
max_speed = 0
for thread in self._search_thread_pool.values():
if thread.speed > max_speed: