[MyPy] Fix mypy for vllm/benchmarks (#39896)

Signed-off-by: Martin Hickey <martin.hickey@ie.ibm.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Martin Hickey
2026-06-23 15:22:29 +00:00
committed by GitHub
co-authored by Harry Mellor
parent 547d2c40d7
commit 275b43183c
24 changed files with 297 additions and 151 deletions
+24 -20
View File
@@ -75,7 +75,7 @@ class RequestFuncInput:
logprobs: int | None = None
extra_headers: dict | None = None
extra_body: dict | None = None
multi_modal_content: dict | list[dict] | None = None
multi_modal_content: dict[str, Any] | list[dict[str, Any]] | None = None
ignore_eos: bool = False
language: str | None = None
request_id: str | None = None
@@ -272,13 +272,13 @@ def _get_chat_content(
request_func_input: RequestFuncInput,
mm_position: Literal["first", "last"] = "last",
) -> list[dict[str, Any]]:
mm_contents = []
mm_contents: list[dict[str, Any]] = []
if request_func_input.multi_modal_content:
mm_content = request_func_input.multi_modal_content
if isinstance(mm_content, list):
mm_contents.extend(request_func_input.multi_modal_content)
mm_contents.extend(mm_content)
elif isinstance(mm_content, dict):
mm_contents.append(request_func_input.multi_modal_content)
mm_contents.append(mm_content)
else:
raise TypeError(
"multi_modal_content must be a dict or list[dict] for openai-chat"
@@ -293,10 +293,11 @@ def _get_chat_content(
for item in prompt
)
):
prompt_dicts: list[dict[str, Any]] = prompt # type: ignore[assignment]
if mm_position == "first":
return mm_contents + prompt
return mm_contents + prompt_dicts
return prompt + mm_contents
return prompt_dicts + mm_contents
text_contents = [{"type": "text", "text": prompt}]
@@ -307,15 +308,15 @@ def _get_chat_content(
def _is_chat_messages(prompt: Any) -> bool:
return (
isinstance(prompt, list)
and prompt
and all(
isinstance(item, dict)
and isinstance(item.get("role"), str)
and isinstance(item.get("content"), (str, list))
for item in prompt
)
if not isinstance(prompt, list):
return False
if not prompt:
return False
return all(
isinstance(item, dict)
and isinstance(item.get("role"), str)
and isinstance(item.get("content"), (str, list))
for item in prompt
)
@@ -325,7 +326,7 @@ def _get_chat_messages(
) -> list[dict[str, Any]]:
prompt = request_func_input.prompt
if _is_chat_messages(prompt):
return prompt
return prompt # type: ignore[return-value]
return [
{
@@ -385,8 +386,8 @@ async def async_request_openai_chat_completions(
if not chunk_bytes:
continue
messages = handler.add_chunk(chunk_bytes)
for message in messages:
message_strings = handler.add_chunk(chunk_bytes)
for message in message_strings:
# NOTE: SSE comments (often used as pings) start with
# a colon. These are not JSON data payload and should
# be skipped.
@@ -790,7 +791,7 @@ async def async_request_infinity_embeddings(
api_url = request_func_input.api_url
_validate_api_url(api_url, "Infinity Embeddings API", "embeddings")
payload = {
payload: dict[str, Any] = {
"model": request_func_input.model_name
if request_func_input.model_name
else request_func_input.model,
@@ -849,7 +850,10 @@ async def async_request_vllm_pooling(
"truncate_prompt_tokens": -1,
}
payload = payload | request_func_input.prompt
if isinstance(request_func_input.prompt, dict):
payload = payload | request_func_input.prompt
else:
payload["input"] = request_func_input.prompt
_update_payload_common(payload, request_func_input)