biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[MISC] cudagraph_capture_sizes related improvements (#26016)

Browse Source 
Signed-off-by: fhl <2410591650@qq.com>
Signed-off-by: fhl2000 <63384265+fhl2000@users.noreply.github.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
fhl2000
2025-10-24 20:11:05 +08:00
committed by GitHub
parent 435be10db9
commit 284cc92275
14 changed files with 303 additions and 110 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
vllm/model_executor/models/config.py
View File

@@ -259,21 +259,19 @@ class GptOssForCausalLMConfig(VerifyAndUpdateConfig):
# Increase the max capture size from 512 to 992 for performance.
# NOTE(woosuk): This will increase the number of CUDA graphs
# from 67 to 81.
scheduler_config = vllm_config.scheduler_config
if len(scheduler_config.cuda_graph_sizes) == 1:
max_capture_size = scheduler_config.cuda_graph_sizes[0]
compilation_config = vllm_config.compilation_config
# Only override when the user has not set either of
# cudagraph_capture_sizes or max_cudagraph_capture_size.
if (
compilation_config.cudagraph_capture_sizes is None
and compilation_config.max_cudagraph_capture_size is None
):
# FIXME(woosuk): When using full cuda graph with FA3, the max
# supported size is 992.
if max_capture_size < 992:
cuda_graph_sizes = [1, 2, 4]
# Step size 8 for small batch sizes
cuda_graph_sizes += [i for i in range(8, 256, 8)]
# Step size 16 for larger batch sizes
cuda_graph_sizes += [i for i in range(256, 993, 16)]
scheduler_config.cuda_graph_sizes = cuda_graph_sizes
logger.info(
"Overriding max cuda graph capture size to %d for performance.", 992
)
compilation_config.max_cudagraph_capture_size = 992
logger.info(
"Overriding max cuda graph capture size to %d for performance.", 992
)
class MambaModelConfig(VerifyAndUpdateConfig):