Skip to content

[AMD][MI35X] 1028 DSV4 sglang mtp agentic #4596

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

Open
1am9trash wants to merge 6 commits into
base: main
Choose a base branch
from
Open

[AMD][MI35X] 1028 DSV4 sglang mtp agentic #4596

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
72c3e63
Update image version and args
1am9trash Oct 28, 2026
6757fb8
Update changelog
1am9trash Oct 28, 2026
d39bc05
Merge branch 'main' into amd/dsv4-mi355x-sglang-1028
1am9trash Oct 28, 2026
56469db
Merge branch 'main' into amd/dsv4-mi355x-sglang-1028
1am9trash Oct 28, 2026
46ff760
Update image
1am9trash Oct 28, 2026
9ab9d0a
Refine description in changelog
1am9trash Oct 28, 2026
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
1 change: 1 addition & 0 deletions ...ncex-e2e/benchmarks/single_node/srt-slurm-recipes/dsv4/sglang/mi355x-fp4-mtp/agentic.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,6 +88,7 @@ override_tp8_c1:
max-running-requests: 2
cuda-graph-max-bs-decode: 2
cuda-graph-backend-prefill: breakable
enable-amd-monokernel: true
benchmark:
env:
CONC: '1'
Expand Down
2 changes: 1 addition & 1 deletion inferencex-e2e/configs/amd-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1537,7 +1537,7 @@ glm5.2-fp4-mi355x-atom-agentic-mtp-mono:
- { tp: 4, kv-offloading: none, conc-list: [1, 2], spec-decoding: mtp, srt-recipe: agentic.yaml }

dsv4-fp4-mi355x-sglang-agentic-mtp:
image: lmsysorg/sglang-rocm:v0.5.22-rocm720-mi35x-20261019
image: lmsysorg/sglang-rocm:v0.5.22-rocm720-mi35x-20261028
model: deepseek-ai/DeepSeek-V4-Pro-0813
model-prefix: dsv4
runner: cluster:mi355x-amds
Expand Down
9 changes: 9 additions & 0 deletions inferencex-e2e/perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11489,3 +11489,12 @@
- "Update the MI355X DeepSeek-V4-Pro-0813 SGLang MoRI disaggregated AgentX serving image from lmsysorg/sglang-rocm:v0.5.22-rocm720-mi35x-20261019 to lmsysorg/sglang-rocm:v0.5.22-rocm10-mi35x-20261026, the latest published MI35x ROCm 10 nightly (digest sha256:edabad83728ab04e3eae39f315eb16261eeb64dbdb051112bd22fc88560b60c4)."
- "Preserve all eight concurrency points (1, 16, 32, 48, 192, 256, 384, 512), prefill/decode topology, DSpark settings, UMBP offload, MoRI transfer, and benchmark client image. The workload binder supplies the serving image to every selected recipe variant."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4571

- config-keys:
- dsv4-fp4-mi355x-sglang-agentic-mtp
scenario-type:
- agentic-coding
description:
- "Bump to lmsysorg/sglang-rocm:v0.5.22-rocm720-mi35x-20261028."
- "Enable --enable-amd-monokernel at concurrency 1. Each DSpark verify there is one request's 7 tokens, which the monokernels run as one persistent launch per layer; steps with more rows run the standard forward."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4596
Loading