Skip to content

[ATOM/ROCm] atom deepseek r1 fp4 mtp3 on mi355x #1028

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
seungrokj wants to merge 2 commits into
base: main
Choose a base branch
from
+12 −4
Open

[ATOM/ROCm] atom deepseek r1 fp4 mtp3 on mi355x #1028

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
283c99f
atom deepseek r1 fp4 mtp3
seungrokj Apr 14, 2026
83f7e74
atom deepseek r1 fp4 mtp3
seungrokj Apr 14, 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
8 changes: 4 additions & 4 deletions .github/configs/amd-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,8 +38,8 @@ dsr1-fp4-mi355x-atom:
- { tp: 8, ep: 1, conc-start: 4, conc-end: 4 }

dsr1-fp4-mi355x-atom-mtp:
image: rocm/atom:rocm7.2.0-ubuntu24.04-pytorch2.9-atom0.1.1
model: amd/DeepSeek-R1-0528-MXFP4
image: rocm/atom:rocm7.2.1-ubuntu24.04-pytorch2.9.1-atom0.1.2.post
model: amd/DeepSeek-R1-0528-MXFP4-MTP-MoEFP4
model-prefix: dsr1
runner: mi355x
precision: fp4
Expand All @@ -51,11 +51,11 @@ dsr1-fp4-mi355x-atom-mtp:
osl: 1024
search-space:
- { tp: 4, conc-start: 4, conc-end: 256, spec-decoding: mtp }
- { tp: 8, conc-start: 4, conc-end: 256, spec-decoding: mtp }
- { tp: 8, conc-start: 4, conc-end: 128, spec-decoding: mtp }
- isl: 8192
osl: 1024
search-space:
#- { tp: 4, conc-start: 32, conc-end: 256, spec-decoding: mtp }
- { tp: 4, conc-start: 4, conc-end: 256, spec-decoding: mtp }
- { tp: 8, conc-start: 4, conc-end: 256, spec-decoding: mtp }

dsr1-fp8-mi300x-sglang:
Expand Down
1 change: 1 addition & 0 deletions benchmarks/single_node/dsr1_fp4_mi355x_atom_mtp.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,6 +55,7 @@ python3 -m atom.entrypoints.openai_server \
-tp $TP \
--kv_cache_dtype fp8 $CALCULATED_MAX_MODEL_LEN $EP \
--method mtp \
--num-speculative-tokens 3 \
> $SERVER_LOG 2>&1 &

SERVER_PID=$!
Expand Down
7 changes: 7 additions & 0 deletions perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1330,3 +1330,10 @@
- "Remove sed hack for TensorRT bug (fixed upstream in v1.3.0rc5)"
- "Remove enable_block_reuse: false from kv_cache_config (default true is now recommended)"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/854

- config-keys:
- dsr1-fp4-mi355x-atom-mtp
description:
- "DeepSeek R1 MI355X FP4 ATOM-MTP config to support MTP 3 tokens"
- "Image: rocm/atom:rocm7.2.1-ubuntu24.04-pytorch2.9.1-atom0.1.2.post"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1028
Loading