apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 21 Packages Projects Releases Wiki Activity

[Feature] support clear data (#3601)

Browse Source 
* [Feature] support clear data

* update

* fix

* fix

* fix

* fix

* fix

* fix

* fix
This commit is contained in:
ltd0924
2025-09-23 10:20:02 +08:00
committed by GitHub
parent 772f0156f3
commit 83720da79f
12 changed files with 85 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastdeploy/entrypoints/openai/serving_completion.py
+11 -1
View File
@@ -227,6 +227,14 @@ class OpenAIServingCompletion:
completion_batched_token_ids = [[] for _ in range(num_choices)]
current_waiting_time = 0
while num_choices > 0:
if self.engine_client.check_model_weight_status():
return ErrorResponse(
error=ErrorInfo(
message="Model weight cleared",
code=ErrorCode.INVALID_VALUE,
type=ErrorType.INVALID_REQUEST_ERROR,
)
)
try:
response = await asyncio.wait_for(response_queue.get(), timeout=10)
current_waiting_time = 0
@@ -281,7 +289,6 @@ class OpenAIServingCompletion:
return res
except Exception as e:
api_server_logger.error(f"Error in completion_full_generator: {e}", exc_info=True)
raise
finally:
self.engine_client.semaphore.release()
if dealer is not None:
@@ -360,6 +367,8 @@ class OpenAIServingCompletion:
)
current_waiting_time = 0
while num_choices > 0:
if self.engine_client.check_model_weight_status():
raise ValueError("Engine is clearing model weight")
try:
response = await asyncio.wait_for(response_queue.get(), timeout=10)
current_waiting_time = 0
@@ -447,6 +456,7 @@ class OpenAIServingCompletion:
choices[-1].finish_reason = self.calc_finish_reason(
request.max_tokens, output_tokens[idx], output, tool_called[idx]
)
send_idx = output.get("send_idx")
# 只有当 send_idx 明确为 0 时才记录日志
if send_idx == 0 and not request.return_token_ids: