Skip to content

Tune MiniMax-M3 AgentX on B200 with updated vLLM / 更新 B200 上的 MiniMax-M3 AgentX 配置 #3435

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
Oseltamivir merged 2 commits into from
Sep 30, 2026
+53 −14
Merged

Tune MiniMax-M3 AgentX on B200 with updated vLLM / 更新 B200 上的 MiniMax-M3 AgentX 配置 #3435

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
7cbda30
perf: tune MiniMax-M3 AgentX on B200
xinli-sw Sep 28, 2026
69178c3
Merge branch 'main' into codex/minimaxm3-b200-b300-vllm-agentx
Oseltamivir Sep 30, 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
57 changes: 44 additions & 13 deletions ...cex-e2e/benchmarks/single_node/srt-slurm-recipes/minimaxm3/vllm/b200-fp4-mtp/agentic.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ base:
name: minimaxm3-fp4-b200-vllm-agentic
model:
path: hf:nvidia/MiniMax-M3-NVFP4
container: vllm/vllm-openai:nightly-1dc464d42681d22f38caf1fdc1eb632dc4421c45
container: vllm/vllm-openai:nightly-af7f9488c2210d67e1033ecdc845b087ee7fe92b
precision: fp4
resources:
gpu_type: b200
Expand All @@ -30,10 +30,11 @@ base:
workers: 1
args:
served-model-name: nvidia/MiniMax-M3-NVFP4
gpu-memory-utilization: 0.9
gpu-memory-utilization: 0.95
block-size: 128
language-model-only: true
enable-prefix-caching: true
enable-chunked-prefill: true
no-enable-flashinfer-autotune: true
reasoning-parser: minimax_m3
tool-call-parser: minimax_m3
Expand All @@ -47,7 +48,7 @@ base:
trust-remote-code: true
# Three-token EAGLE3 with the GQA draft head. Throughput runs switch to
# synthetic rejection at the golden acceptance length.
speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASH_ATTN"}'
speculative-config: '{"method":"eagle3","model":"Inferact/MiniMax-M3-EAGLE3-GQA","num_speculative_tokens":3,"attention_backend":"FLASHINFER","use_local_argmax_reduction":true}'
env:
PYTHONNOUSERSITE: '1'
VLLM_ENGINE_READY_TIMEOUT_S: '3600'
Expand All @@ -61,13 +62,15 @@ base:
AIPERF_REQUIRED_SERVER_METRIC_PREFIX: 'vllm:'

# One variant per point. DRAM points give SimpleCPUOffload the whole host budget
# (TOTAL_CPU_DRAM_GB GiB) with lazy offload.
# (TOTAL_CPU_DRAM_GB decimal GB) with lazy offload.
override_tp8_c1:
roles:
agg:
gpus: 8
args:
tensor-parallel-size: 8
max-num-seqs: 2
max-cudagraph-capture-size: 8
benchmark:
env:
CONC: '1'
Expand All @@ -79,6 +82,8 @@ override_tp4_c1:
gpus: 4
args:
tensor-parallel-size: 4
max-num-seqs: 2
max-cudagraph-capture-size: 8
benchmark:
env:
CONC: '1'
Expand All @@ -90,6 +95,8 @@ override_tp4_c5:
gpus: 4
args:
tensor-parallel-size: 4
max-num-seqs: 10
max-cudagraph-capture-size: 40
benchmark:
env:
CONC: '5'
Expand All @@ -101,6 +108,8 @@ override_tp4_c10:
gpus: 4
args:
tensor-parallel-size: 4
max-num-seqs: 20
max-cudagraph-capture-size: 80
benchmark:
env:
CONC: '10'
Expand All @@ -112,6 +121,8 @@ override_tp4_c15:
gpus: 4
args:
tensor-parallel-size: 4
max-num-seqs: 30
max-cudagraph-capture-size: 120
benchmark:
env:
CONC: '15'
Expand All @@ -123,6 +134,8 @@ override_tp4_c20:
gpus: 4
args:
tensor-parallel-size: 4
max-num-seqs: 40
max-cudagraph-capture-size: 160
benchmark:
env:
CONC: '20'
Expand All @@ -134,7 +147,9 @@ override_tp4_c15_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 30
max-cudagraph-capture-size: 120
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -149,7 +164,9 @@ override_tp4_c20_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 40
max-cudagraph-capture-size: 160
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -164,7 +181,9 @@ override_tp4_c25_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 50
max-cudagraph-capture-size: 200
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -179,7 +198,9 @@ override_tp4_c30_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 60
max-cudagraph-capture-size: 240
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -194,7 +215,9 @@ override_tp4_c32_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 64
max-cudagraph-capture-size: 256
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -209,7 +232,9 @@ override_tp4_c34_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 68
max-cudagraph-capture-size: 272
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -224,7 +249,9 @@ override_tp4_c36_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 72
max-cudagraph-capture-size: 288
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -239,7 +266,9 @@ override_tp4_c38_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 76
max-cudagraph-capture-size: 304
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand All @@ -254,7 +283,9 @@ override_tp4_c40_dram:
gpus: 4
args:
tensor-parallel-size: 4
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1161788653568,"lazy_offload":true}}'
max-num-seqs: 80
max-cudagraph-capture-size: 320
kv-transfer-config: '{"kv_connector":"SimpleCPUOffloadConnector","kv_role":"kv_both","kv_connector_extra_config":{"cpu_bytes_to_use":1082000000000,"lazy_offload":true}}'
env:
VLLM_USE_SIMPLE_KV_OFFLOAD: '1'
benchmark:
Expand Down
2 changes: 1 addition & 1 deletion inferencex-e2e/configs/nvidia-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5674,7 +5674,7 @@ minimaxm3-fp4-b300-trtllm-agentic-mtp:
# the same 3 TB AgentX ceiling before the proportional-GPU rule is applied.
# GPU-resident points receive a zero budget.
minimaxm3-fp4-b200-vllm-agentic-mtp:
image: vllm/vllm-openai:nightly-1dc464d42681d22f38caf1fdc1eb632dc4421c45
image: vllm/vllm-openai:nightly-af7f9488c2210d67e1033ecdc845b087ee7fe92b
model: nvidia/MiniMax-M3-NVFP4
model-prefix: minimaxm3
runner: cluster:b200-nscale
Expand Down
8 changes: 8 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 @@ -9004,6 +9004,14 @@
- "Update B300 vLLM AgentX to DSpark6 on a new image with a sampled concurrency grid and per-mode --kv-cache-memory-bytes pins."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/3477

- config-keys:
- minimaxm3-fp4-b200-vllm-agentic-mtp
scenario-type:
- agentic-coding
description:
- "Update B200 AgentX to vLLM nightly-af7f9488c2210d67e1033ecdc845b087ee7fe92b and enable local argmax reduction for the FlashInfer EAGLE3 draft."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/3435

- config-keys:
- glm5.2-fp4-mi355x-sglang-agentic-mtp
scenario-type:
Expand Down
Loading