修复传入max_tokens=1时的报错 (#3068)

* 修复传入max_tokens=1时的报错

* 修复传入max_tokens=1时的报错

* 修复传入max_tokens=1时的报错

* 修复传入max_tokens=1时的报错

* 修复传入max_tokens=1时的报错

* 修复传入max_tokens=1时的报错
This commit is contained in:
zhuzixuan
2025-07-29 23:49:28 +08:00
committed by GitHub
parent 73cfe1fd37
commit ad7bb52a28
2 changed files with 24 additions and 0 deletions

View File

@@ -344,6 +344,7 @@ class TokenProcessor:
first_token_time=time.time() - task.inference_start_time,
time_in_queue=task.schedule_start_time - task.preprocess_end_time,
preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time,
request_start_time=task.arrival_time,
)
self._record_first_token_metrics(task, current_time)
@@ -476,6 +477,7 @@ class TokenProcessor:
first_token_time=time.time() - task.inference_start_time,
time_in_queue=task.schedule_start_time - task.preprocess_end_time,
preprocess_cost_time=task.preprocess_end_time - task.preprocess_start_time,
request_start_time=task.arrival_time,
)
self._record_first_token_metrics(task, current_time)