-
Notifications
You must be signed in to change notification settings - Fork 181
[NV] Add MiniMax-M2.5 FP8 GB300 Dynamo vLLM recipes #1647
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.
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
jasonlizhengjian
wants to merge
7
commits into
main
Choose a base branch
from
nv/jasonli/minimaxm2.5-fp8-gb300-dynamo-vllm
base: main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Changes from all commits
Commits
Show all changes
7 commits
Select commit
Hold shift + click to select a range
af6d92f
Add MiniMax-M2.5 FP8 GB300 Dynamo vLLM recipes
jasonlizhengjian a72f7c7
Update GB300 FP8 MiniMax changelog PR link
jasonlizhengjian 4edbeea
Fix GB300 eval artifact copy
jasonlizhengjian cc720f0
Handle existing GB300 eval artifacts
jasonlizhengjian e00c5ec
Merge main into GB300 FP8 MiniMax PR
jasonlizhengjian 8f59225
Do not fail GB300 eval artifact copy
jasonlizhengjian f72628c
Merge latest main into GB300 FP8 MiniMax PR
jasonlizhengjian File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
64 changes: 64 additions & 0 deletions
64
...hmarks/multi_node/srt-slurm-recipes/vllm/minimax-m2.5-fp8/1k1k/disagg-gb300-1p1d-tp4.yaml
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,64 @@ | ||
| name: "minimax-m2.5-vllm-disagg-gb300-1p1d-tp4" | ||
|
|
||
| model: | ||
| path: "minimax-m2.5-fp8" | ||
| container: "vllm/vllm-openai:v0.20.1" | ||
| precision: "fp8" | ||
|
|
||
| dynamo: | ||
| install: true | ||
| wheel: "1.2.0.dev20260526" | ||
|
|
||
| setup_script: install-deps.sh | ||
|
|
||
| resources: | ||
| gpu_type: "gb300" | ||
| gpus_per_node: 4 | ||
| prefill_nodes: 1 | ||
| decode_nodes: 1 | ||
| prefill_workers: 1 | ||
| decode_workers: 1 | ||
| gpus_per_prefill: 1 | ||
| gpus_per_decode: 4 | ||
|
|
||
| frontend: | ||
| type: dynamo | ||
| enable_multiple_frontends: false | ||
|
|
||
| backend: | ||
| type: vllm | ||
| connector: null | ||
|
|
||
| prefill_environment: | ||
| VLLM_ENGINE_READY_TIMEOUT_S: "3600" | ||
| VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl" | ||
|
|
||
| decode_environment: | ||
| VLLM_ENGINE_READY_TIMEOUT_S: "3600" | ||
| VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl" | ||
|
|
||
| vllm_config: | ||
| prefill: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 1 | ||
| pipeline-parallel-size: 1 | ||
| safetensors-load-strategy: "prefetch" | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| stream-interval: 32 | ||
|
|
||
| decode: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 4 | ||
| safetensors-load-strategy: "prefetch" | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| stream-interval: 32 | ||
|
|
||
| benchmark: | ||
| type: "sa-bench" | ||
| isl: 1024 | ||
| osl: 1024 | ||
| concurrencies: "8x16x32x64x128" | ||
69 changes: 69 additions & 0 deletions
69
...hmarks/multi_node/srt-slurm-recipes/vllm/minimax-m2.5-fp8/1k1k/disagg-gb300-1p2d-tp4.yaml
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,69 @@ | ||
| name: "minimax-m2.5-vllm-disagg-gb300-1p2d-tp4" | ||
|
|
||
| model: | ||
| path: "minimax-m2.5-fp8" | ||
| container: "vllm/vllm-openai:v0.20.1" | ||
| precision: "fp8" | ||
|
|
||
| dynamo: | ||
| install: true | ||
| wheel: "1.2.0.dev20260526" | ||
|
|
||
|
|
||
| setup_script: install-deps.sh | ||
|
|
||
| resources: | ||
| gpu_type: "gb300" | ||
| gpus_per_node: 4 | ||
| prefill_nodes: 1 | ||
| decode_nodes: 2 | ||
| prefill_workers: 1 | ||
| decode_workers: 2 | ||
| gpus_per_prefill: 1 | ||
| gpus_per_decode: 4 | ||
|
|
||
| frontend: | ||
| type: dynamo | ||
| enable_multiple_frontends: false | ||
|
|
||
| backend: | ||
| type: vllm | ||
| connector: null | ||
|
|
||
| prefill_environment: | ||
| VLLM_ENGINE_READY_TIMEOUT_S: "3600" | ||
| VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl" | ||
|
|
||
| decode_environment: | ||
| VLLM_ENGINE_READY_TIMEOUT_S: "3600" | ||
| VLLM_FLASHINFER_ALLREDUCE_BACKEND: "mnnvl" | ||
|
|
||
| vllm_config: | ||
| prefill: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 1 | ||
| pipeline-parallel-size: 1 | ||
| safetensors-load-strategy: "prefetch" | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| stream-interval: 32 | ||
|
|
||
| decode: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 4 | ||
| safetensors-load-strategy: "prefetch" | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| stream-interval: 32 | ||
|
|
||
| benchmark: | ||
| type: "sa-bench" | ||
| isl: 1024 | ||
| osl: 1024 | ||
| concurrencies: "32x64x128x256x512" | ||
| # warmup_prompts: 1 | ||
| # use_chat_template: false | ||
| # req_rate: "inf" | ||
| # random_range_ratio: 1.0 |
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Missing GB300 slurm limits
Medium Severity
New GB300 MiniMax Dynamo vLLM recipes jump straight from
setup_scripttoresourcesand never setslurm.time_limitorhealth_check, unlike existing GB300 vLLM disagg recipes on the same runner. Jobs therefore keep the launcher’s four-hour default while engine startup allows up to an hour, so long multi-concurrency sweeps can be killed before benchmarks finish.Additional Locations (1)
benchmarks/multi_node/srt-slurm-recipes/vllm/minimax-m2.5-fp8/1k1k/disagg-gb300-2p2d-dep4-hi-conc.yaml#L22-L33Reviewed by Cursor Bugbot for commit cc720f0. Configure here.