phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-22 18:46:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

chore: add mypy inference parallel utils

Browse source 
Signed-off-by: Mustafa Elbehery <melbeher@redhat.com>
This commit is contained in:
Mustafa Elbehery 2025-07-09 00:01:20 +02:00
parent 51b179e1c5
commit dc043fa395
2 changed files with 3 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

6
llama_stack/providers/inline/inference/meta_reference/parallel_utils.py
View file

@ -98,7 +98,7 @@ class ProcessingMessageWrapper(BaseModel):
def mp_rank_0() -> bool:
return get_model_parallel_rank() == 0
return bool(get_model_parallel_rank() == 0)
def encode_msg(msg: ProcessingMessage) -> bytes:
@ -125,7 +125,7 @@ def retrieve_requests(reply_socket_url: str):
reply_socket.send_multipart([client_id, encode_msg(obj)])
while True:
tasks = [None]
tasks: list[ProcessingMessage | None] = [None]
if mp_rank_0():
client_id, maybe_task_json = maybe_get_work(reply_socket)
if maybe_task_json is not None:
@ -152,7 +152,7 @@ def retrieve_requests(reply_socket_url: str):
break
for obj in out:
updates = [None]
updates: list[ProcessingMessage | None] = [None]
if mp_rank_0():
_, update_json = maybe_get_work(reply_socket)
update = maybe_parse_message(update_json)

1
pyproject.toml
View file

@ -255,7 +255,6 @@ exclude = [
"^llama_stack/models/llama/llama3/generation\\.py$",
"^llama_stack/models/llama/llama3/multimodal/model\\.py$",
"^llama_stack/models/llama/llama4/",
"^llama_stack/providers/inline/inference/meta_reference/parallel_utils\\.py$",
"^llama_stack/providers/inline/inference/meta_reference/quantization/fp8_impls\\.py$",
"^llama_stack/providers/inline/inference/sentence_transformers/sentence_transformers\\.py$",
"^llama_stack/providers/inline/inference/vllm/",