apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-07 01:22:59 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Bug fix] Fix perf in mixed deployment with yiyan adpater (#3703)

Browse Source 
Co-authored-by: YuBaoku <49938469+EmmonsCurse@users.noreply.github.com>
This commit is contained in:
chenjian
2025-09-01 14:06:09 +08:00
committed by GitHub
parent a424ab907f
commit c49c43d51c
2 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
fastdeploy/output/token_processor.py
View File

@@ -360,6 +360,7 @@ class TokenProcessor:
metrics = RequestMetrics( metrics = RequestMetrics(
arrival_time=task.arrival_time, arrival_time=task.arrival_time,
inference_start_time=task.inference_start_time, inference_start_time=task.inference_start_time,
model_execute_time=time.time() - task.inference_start_time,
first_token_time=time.time() - task.inference_start_time, first_token_time=time.time() - task.inference_start_time,
time_in_queue=task.schedule_start_time - task.preprocess_end_time, time_in_queue=task.schedule_start_time - task.preprocess_end_time,
preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time, preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time,
@@ -503,6 +504,7 @@ class TokenProcessor:
metrics = RequestMetrics( metrics = RequestMetrics(
arrival_time=task.arrival_time, arrival_time=task.arrival_time,
inference_start_time=task.inference_start_time, inference_start_time=task.inference_start_time,
model_execute_time=time.time() - task.inference_start_time,
first_token_time=time.time() - task.inference_start_time, first_token_time=time.time() - task.inference_start_time,
time_in_queue=task.schedule_start_time - task.preprocess_end_time, time_in_queue=task.schedule_start_time - task.preprocess_end_time,
preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time, preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time,
@@ -514,6 +516,7 @@ class TokenProcessor:
else: else:
metrics = RequestMetrics( metrics = RequestMetrics(
arrival_time=time.time(), arrival_time=time.time(),
model_execute_time=time.time() - task.inference_start_time,
request_start_time=task.arrival_time, request_start_time=task.arrival_time,
) )
self.number_of_output_tokens += len(token_ids) self.number_of_output_tokens += len(token_ids)

2
fastdeploy/scheduler/local_scheduler.py
View File

@@ -209,7 +209,7 @@ class LocalScheduler:
return (token_num + block_size - 1) // block_size return (token_num + block_size - 1) // block_size
def get_unhandled_request_num(self): def get_unhandled_request_num(self):
return len(self.requests) return len(self.ids) - self.ids_read_cursor
def get_requests( def get_requests(
self, self,