Skip to content

[Bugfix] Miscalculated latency lead to time_to_first_token_seconds inaccurate. #6686

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Jul 24, 2024
Merged

[Bugfix] Miscalculated latency lead to time_to_first_token_seconds inaccurate. #6686

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
4dcf52f
[bugfix] do_log_stats() Must call prometheus's logger first.
AllenDou Jul 23, 2024
1e6c512
need to call _get_stats() only once.
AllenDou Jul 23, 2024
995d3b5
Update vllm/engine/llm_engine.py
Yard1 Jul 23, 2024
7c9b78d
typo fix.
AllenDou Jul 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion vllm/engine/llm_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -949,8 +949,9 @@ def do_log_stats(
model_output: Optional[List[SamplerOutput]] = None) -> None:
"""Forced log when no requests active."""
if self.log_stats:
stats = self._get_stats(scheduler_outputs, model_output)
for logger in self.stat_loggers.values():
logger.log(self._get_stats(scheduler_outputs, model_output))
logger.log(stats)

def _get_stats(
self,
Expand Down
2 changes: 1 addition & 1 deletion vllm/spec_decode/spec_decode_worker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -484,7 +484,7 @@ def _run_non_driver_rank(self) -> bool:
for both speculation cases (num_lookahead_slots>0) and non-speculation
cases (e.g. prefill).

Returns True iff there are remaining sequences to process.
Returns True if there are remaining sequences to process.
"""
assert self.rank != self._driver_rank

Expand Down