Skip to content
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

[Core] Update outlines and increase its threadpool size #11140

Merged
merged 3 commits into from
Dec 14, 2024
Merged

[Core] Update outlines and increase its threadpool size #11140

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
1a45ba3
[Build/CI] Update outlines to 0.10.0
russellb Dec 12, 2024
0095c6a
outlines: Scale thread pool size based on cpu count
russellb Dec 12, 2024
f364718
Update to outlines-0.1.11
russellb Dec 13, 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
2 changes: 1 addition & 1 deletion requirements-common.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@ prometheus_client >= 0.18.0
prometheus-fastapi-instrumentator >= 7.0.0
tiktoken >= 0.6.0 # Required for DBRX tokenizer
lm-format-enforcer >= 0.10.9, < 0.11
outlines == 0.1.9
outlines == 0.1.11
xgrammar >= 0.1.6; platform_machine == "x86_64"
typing_extensions >= 4.10
filelock >= 3.16.1 # need to contain https://github.com/tox-dev/filelock/pull/317
Expand Down
11 changes: 10 additions & 1 deletion vllm/model_executor/guided_decoding/outlines_decoding.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
import asyncio
import concurrent.futures
import os
from enum import Enum
from json import dumps as json_dumps
from re import escape as regex_escape
Expand Down Expand Up @@ -48,6 +49,11 @@ class GuidedDecodingMode(Enum):

global_thread_pool = None # used for generating logits processor fsm

# It's not yet clear that using more provides a benefit, and it could
# potentially starve other processes on the machine. We'll cap this for now and
# adjust later if testing proves it to help overcome a bottleneck.
_MAX_THREADPOOL_WORKERS = 16


async def get_outlines_guided_decoding_logits_processor(
guided_params: GuidedDecodingParams, tokenizer: PreTrainedTokenizerBase
Expand All @@ -65,8 +71,11 @@ async def get_outlines_guided_decoding_logits_processor(
return None

if global_thread_pool is None:
max_workers = os.cpu_count() or 2
mgoin marked this conversation as resolved.
Show resolved Hide resolved
if max_workers > _MAX_THREADPOOL_WORKERS:
max_workers = _MAX_THREADPOOL_WORKERS
global_thread_pool = concurrent.futures.ThreadPoolExecutor(
max_workers=2)
max_workers=max_workers)
loop = asyncio.get_running_loop()

return await loop.run_in_executor(global_thread_pool,
Expand Down
Loading