Skip to content

[bugfix] Add 'disaggregation_mode' parameter to warmup function when compile deep_gemm manually #8618

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
zhyncs merged 4 commits into from
Aug 1, 2025
+8 −1
Merged

[bugfix] Add 'disaggregation_mode' parameter to warmup function when compile deep_gemm manually #8618

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
dccda4a
Add 'disaggregation_mode' parameter to warmup function when compile d…
lbh2001 Jul 31, 2025
a992a05
Merge branch 'main' into fix_compile_deep_gemm
lbh2001 Aug 1, 2025
83457d9
Merge branch 'main' into fix_compile_deep_gemm
zhyncs Aug 1, 2025
d0f4557
upd
zhyncs Aug 1, 2025
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: 6 additions & 1 deletion python/sglang/compile_deep_gemm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@

import requests

from sglang.srt.disaggregation.utils import FAKE_BOOTSTRAP_HOST
Comment thread
lbh2001 marked this conversation as resolved.
from sglang.srt.entrypoints.http_server import launch_server
from sglang.srt.managers.io_struct import GenerateReqInput
from sglang.srt.managers.tokenizer_manager import TokenizerManager
Expand Down Expand Up @@ -52,7 +53,7 @@ def from_cli_args(cls, args: argparse.Namespace):


@warmup("compile-deep-gemm")
async def warm_up_compile(tokenizer_manager: TokenizerManager):
async def warm_up_compile(disaggregation_mode: str, tokenizer_manager: TokenizerManager):
print("\nGenerate warm up request for compiling DeepGEMM...\n")
generate_req_input = GenerateReqInput(
input_ids=[0, 1, 2, 3],
Expand All @@ -62,6 +63,10 @@ async def warm_up_compile(tokenizer_manager: TokenizerManager):
"ignore_eos": True,
},
)
if disaggregation_mode != "null":
Comment thread
lbh2001 marked this conversation as resolved.
generate_req_input.bootstrap_room = 0
generate_req_input.bootstrap_host = FAKE_BOOTSTRAP_HOST

await tokenizer_manager.generate_request(generate_req_input, None).__anext__()


Expand Down