xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-04-11 06:57:03 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Perf] Use small max_num_batched_tokens for A100 (#17885)

Browse Source 
Signed-off-by: KuntaiDu <kuntai@uchicago.edu>
This commit is contained in:
Kuntai Du 2025-05-11 00:53:23 -07:00 committed by GitHub
parent 90d0a74b60
commit 9112155283
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/engine/arg_utils.py
View File

@ -1438,11 +1438,15 @@ class EngineArgs:
from vllm.platforms import current_platform
try:
device_memory = current_platform.get_device_total_memory()
device_name = current_platform.get_device_name().lower()
except Exception:
# This is only used to set default_max_num_batched_tokens
device_memory = 0
if device_memory >= 70 * GiB_bytes:
# NOTE(Kuntai): Setting large `max_num_batched_tokens` for A100 reduces
# throughput, see PR #17885 for more details.
# So here we do an extra device name check to prevent such regression.
if device_memory >= 70 * GiB_bytes and "a100" not in device_name:
# For GPUs like H100 and MI300x, use larger default values.
default_max_num_batched_tokens = {
UsageContext.LLM_CLASS: 16384,