apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-06 00:57:33 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Fix] mv connection_manager init (#3902)
Some checks failed
CE Compile Job / ce_job_pre_check (push) Has been cancelled
Details
CE Compile Job / print_ce_job_pre_check_outputs (push) Has been cancelled
Details
CE Compile Job / FD-Clone-Linux (push) Has been cancelled
Details
CE Compile Job / Show Code Archive Output (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8090 (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8689 (push) Has been cancelled
Details
CE Compile Job / CE_UPLOAD (push) Has been cancelled
Details

Browse Source 
* Update serving_chat.py

* Update serving_completion.py

* Update serving_completion.py

* mv connection_manager init

---------

Co-authored-by: Yuanle Liu <yuanlehome@163.com>
This commit is contained in:
ltd0924
2025-09-05 17:42:36 +08:00
committed by GitHub
parent 199f88ce1e
commit 173e4df982
3 changed files with 1 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
fastdeploy/entrypoints/openai/api_server.py
View File

@@ -182,6 +182,7 @@ async def lifespan(app: FastAPI):
workers=args.workers,
tool_parser=args.tool_call_parser,
)
await engine_client.connection_manager.initialize()
app.state.dynamic_load_weight = args.dynamic_load_weight
model_handler = OpenAIServingModels(
model_paths,

8
fastdeploy/entrypoints/openai/serving_chat.py
View File

@@ -74,12 +74,6 @@ class OpenAIServingChat:
self.master_ip = "0.0.0.0"
api_server_logger.info(f"master ip: {self.master_ip}")
async def _ensure_connection_manager(self):
"""ensure connection manager initialized"""
if not self.engine_client.connection_initialized:
await self.engine_client.connection_manager.initialize()
self.engine_client.connection_initialized = True
def _check_master(self):
return self.engine_client.is_master
@@ -206,7 +200,6 @@ class OpenAIServingChat:
api_server_logger.info(f"create chat completion request: {request_id}")
try:
await self._ensure_connection_manager()
dealer, response_queue = await self.engine_client.connection_manager.get_connection(request_id)
dealer.write([b"", request_id.encode("utf-8")])
choices = []
@@ -419,7 +412,6 @@ class OpenAIServingChat:
include_stop_str_in_output = request.include_stop_str_in_output
try:
await self._ensure_connection_manager()
dealer, response_queue = await self.engine_client.connection_manager.get_connection(request_id)
dealer.write([b"", request_id.encode("utf-8")])
final_res = None

8
fastdeploy/entrypoints/openai/serving_completion.py
View File

@@ -51,12 +51,6 @@ class OpenAIServingCompletion:
else:
self.master_ip = "0.0.0.0"
async def _ensure_connection_manager(self):
"""ensure connection manager initialized"""
if not self.engine_client.connection_initialized:
await self.engine_client.connection_manager.initialize()
self.engine_client.connection_initialized = True
def _check_master(self):
return self.engine_client.is_master
@@ -208,7 +202,6 @@ class OpenAIServingCompletion:
try:
request_ids = [f"{request_id}-{i}" for i in range(num_choices)]
# create dealer
await self._ensure_connection_manager()
dealer, response_queue = await self.engine_client.connection_manager.get_connection(
request_id, num_choices
)
@@ -314,7 +307,6 @@ class OpenAIServingCompletion:
Process the stream completion request.
"""
try:
await self._ensure_connection_manager()
dealer, response_queue = await self.engine_client.connection_manager.get_connection(
request_id, num_choices
)