Use full search range for every trial

src/aituner/store.py

@@ -85,15 +85,11 @@ class StudyStore:
         trial_root = self.study_root(study.study_id) / "trials" / trial_id
         trial_root.mkdir(parents=True, exist_ok=True)
         parallel_size = _parallel_size_for_proposal(study=study, proposal=proposal)
-        search_low = _derive_search_floor(study=study, state=state, parallel_size=parallel_size)
         spec = TrialSpec(
             study_id=study.study_id,
             trial_id=trial_id,
             config_patch=proposal.config_patch,
-            search=replace(
-                study.search,
-                low=search_low,
-            ),
+            search=study.search,
             study_spec_path=str((self.study_root(study.study_id) / "study_spec.source").resolve()),
             artifact_dir=str(trial_root),
             probe_log_path=str(trial_root / "probe_history.json"),
@@ -305,15 +301,3 @@ def _request_rate_per_gpu(best_request_rate: Any, parallel_size: int | None) ->
         return None
     return float(best_request_rate) / float(parallel_size)
 
-
-def _derive_search_floor(*, study: StudySpec, state: StudyState, parallel_size: int) -> float:
-    low = study.search.low
-    high = study.search.high
-    group_incumbent = (state.best_by_parallel_size or {}).get(str(parallel_size))
-    if isinstance(group_incumbent, dict) and isinstance(
-        group_incumbent.get("best_sampling_u"), (int, float)
-    ):
-        candidate = float(group_incumbent["best_sampling_u"])
-    else:
-        candidate = low
-    return min(high, max(low, candidate))