scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-28 01:32:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 43

fix: address bugs in async pass-through, anthropic cache token detection, rerank tests

Browse source 
- async_get_available_deployment_for_pass_through: enforce blocked check on specific deployments
- cost_calculator: detect anthropic-style usage by attribute presence (not truthiness) to avoid mixing OpenAI cached_tokens into anthropic normalization when read=0
- dashscope rerank tests: pass request to httpx.Response constructions for consistency

Co-authored-by: Yassin Kortam <yassin@berri.ai>
This commit is contained in:
Cursor Agent 2026-05-18 14:07:39 +00:00
parent 9de0fcea93
commit d89de671d8
No known key found for this signature in database
3 changed files with 17 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
litellm/cost_calculator.py
View file

@ -375,11 +375,11 @@ def cost_per_token( # noqa: PLR0915
_anthropic_read = getattr(usage_object, "cache_read_input_tokens", None)
_anthropic_create = getattr(usage_object, "cache_creation_input_tokens", None)
if _anthropic_read or _anthropic_create:
if _anthropic_read is not None or _anthropic_create is not None:
_is_anthropic_style = True
if _anthropic_read:
if _anthropic_read is not None:
_cache_read_tokens = float(_anthropic_read)
if _anthropic_create:
if _anthropic_create is not None:
_cache_creation_tokens = float(_anthropic_create)
if not _cache_read_tokens and cache_read_input_tokens:

6
litellm/router.py
View file

@ -10426,6 +10426,12 @@ class Router:
# 3. If specific deployment returned, verify if it supports pass-through
if isinstance(healthy_deployments, dict):
if (healthy_deployments.get("model_info") or {}).get("blocked") is True:
raise litellm.ServiceUnavailableError(
message=f"Model '{model}' is administratively paused. Contact your proxy admin to unblock it.",
model=model,
llm_provider="",
)
litellm_params = healthy_deployments.get("litellm_params", {})
if litellm_params.get("use_in_pass_through"):
return healthy_deployments

10
tests/test_litellm/llms/dashscope/test_dashscope_rerank_transformation.py
View file

@ -166,7 +166,9 @@ class TestDashScopeRerankResponse:
def _resp(self, body, status_code=200):
return httpx.Response(
status_code=status_code, content=json.dumps(body).encode()
status_code=status_code,
content=json.dumps(body).encode(),
request=httpx.Request("POST", "https://example.com"),
)
def test_success_response(self):
@ -291,7 +293,11 @@ class TestDashScopeRerankResponse:
assert "Invalid API-key provided." in str(exc_info.value)
def test_non_json_response_raises(self):
bad = httpx.Response(status_code=500, content=b"<html>bad gateway</html>")
bad = httpx.Response(
status_code=500,
content=b"<html>bad gateway</html>",
request=httpx.Request("POST", "https://example.com"),
)
with pytest.raises(DashScopeError):
self.config.transform_rerank_response(
model="qwen3-rerank",