Skip to content

[Bugfix] Set num_cached_tokens = num_computed_tokens if unset #1471

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

Closed
gcanlin wants to merge 1 commit into from
Closed

[Bugfix] Set num_cached_tokens = num_computed_tokens if unset #1471

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
7 changes: 7 additions & 0 deletions vllm_omni/core/sched/omni_generation_scheduler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -107,6 +107,9 @@ def schedule(self) -> SchedulerOutput:
num_scheduled_tokens[request.request_id] = num_new_tokens
cached_prompt_token_ids[request.request_id] = request.prompt_token_ids
token_budget -= num_new_tokens
# Set num_cached_tokens to match base scheduler behavior for metrics
if request.num_cached_tokens < 0:
request.num_cached_tokens = num_computed_tokens
scheduled_running_reqs.append(request)
req_index += 1

Expand Down Expand Up @@ -168,6 +171,10 @@ def schedule(self) -> SchedulerOutput:
req_to_new_blocks[request.request_id] = new_blocks
num_scheduled_tokens[request.request_id] = num_new_tokens
token_budget -= num_new_tokens
# Set num_cached_tokens to match base scheduler behavior for metrics
# New requests have 0 computed tokens at this point
if request.num_cached_tokens < 0:
request.num_cached_tokens = 0
scheduled_new_reqs.append(request)

# Return skipped waiting requests
Expand Down