perf(cli): cap /model picker custom-endpoint probe at 1.5s

The interactive /model picker probes the current custom endpoint live via
fetch_api_models(), which defaults to a 5s timeout. A slow or flaky custom
endpoint blocks the picker for up to 5s on open. The lmstudio picker probe
already uses a 1.5s timeout for exactly this reason; apply the same fast-fail
bound to the three custom-provider probe sites, gated on for_picker so the
non-picker (5s) path is unchanged.
This commit is contained in:
ZachariahChu
2026-07-26 20:29:45 +08:00
committed by kshitij
parent 4ea379ca2e
commit de0ce24c2e
+7 -1
View File
@@ -2702,6 +2702,7 @@ def list_authenticated_providers(
live_models = fetch_api_models(
api_key,
api_url,
timeout=1.5 if for_picker else 5.0, # picker: fail fast so a slow custom endpoint doesn't block /model
headers=_extra_headers_from_config(ep_cfg) or None,
)
if live_models:
@@ -2769,7 +2770,11 @@ def list_authenticated_providers(
try:
from hermes_cli.models import fetch_api_models
_live_models = fetch_api_models("", str(current_base_url).strip().rstrip("/"))
_live_models = fetch_api_models(
"",
str(current_base_url).strip().rstrip("/"),
timeout=1.5 if for_picker else 5.0, # picker: fail fast on a slow current endpoint
)
if _live_models:
_models = _live_models
except Exception:
@@ -3012,6 +3017,7 @@ def list_authenticated_providers(
live_models = fetch_api_models(
api_key,
api_url,
timeout=1.5 if for_picker else 5.0, # picker: fail fast so a slow custom endpoint doesn't block /model
headers=grp.get("extra_headers") or None,
)
if live_models: