-
Notifications
You must be signed in to change notification settings - Fork 156
[NV] Add GB200 MegaMOE max throughput recipe #1218
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
Merged
Merged
Changes from all commits
Commits
Show all changes
2 commits
Select commit
Hold shift + click to select a range
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
Some comments aren't visible on the classic Files Changed page.
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
150 changes: 150 additions & 0 deletions
150
...rks/multi_node/srt-slurm-recipes/vllm/deepseek-v4/8k1k/disagg-gb200-low-middle-curve.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 +1,150 @@ | ||
| name: "svf-vllm-disagg-gb200-low-middle-curve" | ||
|
|
||
| # Mirrored from NVIDIA/srt-slurm aflowers/vllm-gb200-v0.20.0 branch: | ||
| # recipes/vllm/deepseek-v4-pro/GB200/8k1k/disagg-gb200-low-middle-curve.yaml | ||
| # | ||
| # Topology: 1 prefill (DEP=8) + 4 decode (TP=8). 11 nodes total with a | ||
| # dedicated NATS/etcd infra node. Low-middle curve points at concurrencies | ||
| # 256 and 512. | ||
| # | ||
| # Local deltas vs upstream: | ||
| # * model.path alias renamed deepseekv4-fp4 -> deepseek-v4-pro to match | ||
| # SRT_SLURM_MODEL_PREFIX in runners/launch_gb200-nv.sh. | ||
| # * model.container set to vllm/vllm-openai:v0.20.0-ubuntu2404 to | ||
| # match nvidia-master.yaml image (which the launch script registers as | ||
| # the alias key in srtslurm.yaml). Upstream variants ship either the | ||
| # non-dynamo floating tag or a sha256 pin. | ||
| # * slurm.time_limit + health_check set to 8h / 1440 attempts to | ||
| # absorb cold-cache /mnt/numa1 model loads. | ||
| model: | ||
| path: "deepseek-v4-pro" | ||
| container: "vllm/vllm-openai:v0.20.0-ubuntu2404" | ||
| precision: "fp4" | ||
|
|
||
| dynamo: | ||
| install: true | ||
| wheel: "1.2.0.dev20260426" | ||
|
|
||
| setup_script: vllm-container-deps.sh | ||
|
|
||
| slurm: | ||
| time_limit: "8:00:00" | ||
|
|
||
| health_check: | ||
| max_attempts: 1440 | ||
| interval_seconds: 10 | ||
| resources: | ||
| gpu_type: "gb200" | ||
| gpus_per_node: 4 | ||
| prefill_nodes: 2 | ||
| decode_nodes: 8 | ||
| prefill_workers: 1 | ||
| decode_workers: 4 | ||
| gpus_per_prefill: 8 | ||
| gpus_per_decode: 8 | ||
|
|
||
| infra: | ||
| etcd_nats_dedicated_node: true | ||
|
|
||
| frontend: | ||
| type: dynamo | ||
| enable_multiple_frontends: false | ||
| backend: | ||
| type: vllm | ||
| connector: null | ||
| prefill_environment: | ||
| TILELANG_CLEANUP_TEMP_FILES: "1" | ||
| VLLM_USE_NCCL_SYMM_MEM: "1" | ||
| NCCL_CUMEM_ENABLE: "1" | ||
| NCCL_MNNVL_ENABLE: "1" | ||
| NCCL_NVLS_ENABLE: "1" | ||
| VLLM_SERVER_DEV_MODE: "1" | ||
| VLLM_SPARSE_INDEXER_MAX_LOGITS_MB: "1024" | ||
| VLLM_MAX_TOKENS_PER_EXPERT_FP4_MOE: "2048" | ||
| # VLLM_RANDOMIZE_DP_DUMMY_INPUTS: "1" | ||
| # VLLM_MOE_ROUTING_SIMULATION_STRATEGY: "uniform_random" | ||
| UCX_MEMTYPE_CACHE: "n" | ||
| UCX_MEMTYPE_REG_WHOLE: "n" | ||
| UCX_TLS: "cuda_copy,cuda_ipc,tcp" | ||
| UCX_CUDA_IPC_ENABLE_MNNVL: "y" | ||
| NCCL_P2P_LEVEL: NVL | ||
| decode_environment: | ||
| TILELANG_CLEANUP_TEMP_FILES: "1" | ||
| VLLM_USE_NCCL_SYMM_MEM: "1" | ||
| NCCL_CUMEM_ENABLE: "1" | ||
| NCCL_MNNVL_ENABLE: "1" | ||
| NCCL_NVLS_ENABLE: "1" | ||
| VLLM_SERVER_DEV_MODE: "1" | ||
| # VLLM_RANDOMIZE_DP_DUMMY_INPUTS: "1" | ||
| # VLLM_MOE_ROUTING_SIMULATION_STRATEGY: "uniform_random" | ||
| UCX_MEMTYPE_CACHE: "n" | ||
| UCX_MEMTYPE_REG_WHOLE: "n" | ||
| UCX_TLS: "cuda_copy,cuda_ipc,tcp" | ||
| UCX_CUDA_IPC_ENABLE_MNNVL: "y" | ||
| NCCL_P2P_LEVEL: NVL | ||
| vllm_config: | ||
| prefill: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| served-model-name: "deepseek-ai/DeepSeek-V4-Pro" | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 1 | ||
| pipeline-parallel-size: 1 | ||
| data-parallel-size: 8 | ||
| data-parallel-rpc-port: 13345 | ||
| enable-expert-parallel: true | ||
| enable-ep-weight-filter: true | ||
| enforce-eager: true | ||
| max-model-len: 16384 | ||
| max-num-seqs: 16 | ||
| max-num-batched-tokens: 32768 | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| no-enable-flashinfer-autotune: true | ||
| no-async-scheduling: true | ||
| block-size: 256 | ||
| gpu-memory-utilization: 0.8 | ||
| no-disable-hybrid-kv-cache-manager: true | ||
| enable-sleep-mode: true | ||
| numa-bind: true | ||
| offload-group-size: 3 | ||
| offload-num-in-group: 1 | ||
| offload-prefetch-step: 2 | ||
| # offload-params: "w13_weight w2_weight w13_weight_scale w2_weight_scale wq_b wo_a wo_b shared_experts" | ||
| tokenizer-mode: deepseek_v4 | ||
| decode: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| served-model-name: "deepseek-ai/DeepSeek-V4-Pro" | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 8 | ||
| pipeline-parallel-size: 1 | ||
| # data-parallel-size: 8 | ||
| # data-parallel-rpc-port: 13345 | ||
| # enable-expert-parallel: true | ||
| max-model-len: 16384 | ||
| max-num-seqs: 256 | ||
| max-cudagraph-capture-size: 256 | ||
| max-num-batched-tokens: 256 | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| block-size: 256 | ||
| compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY","mode":0}' | ||
| gpu-memory-utilization: 0.9 | ||
| stream-interval: 50 | ||
| no-disable-hybrid-kv-cache-manager: true | ||
| enable-sleep-mode: true | ||
| tokenizer-mode: deepseek_v4 | ||
| benchmark: | ||
| type: "sa-bench" | ||
| isl: 8192 | ||
| osl: 1024 | ||
| concurrencies: "256x512" | ||
| req_rate: "inf" | ||
| use_chat_template: true | ||
| custom_tokenizer: "sa_bench_tokenizers.vllm_deepseek_v4.VLLMDeepseekV4Tokenizer" | ||
|
|
||
| identity: | ||
| container: | ||
| image: "vllm/vllm-openai:v0.20.0-ubuntu2404" | ||
| frameworks: | ||
| dynamo: "1.2.0.dev20260426" | ||
| vllm: "0.20.0" |
154 changes: 154 additions & 0 deletions
154
...arks/multi_node/srt-slurm-recipes/vllm/deepseek-v4/8k1k/disagg-gb200-max-tpt-megamoe.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 +1,154 @@ | ||
| name: "svf-vllm-disagg-gb200-max-tpt-megamoe" | ||
|
|
||
| # Mirrored from NVIDIA/srt-slurm aflowers/vllm-gb200-v0.20.0 branch: | ||
| # recipes/vllm/deepseek-v4-pro/GB200/8k1k/disagg-gb200-max-tpt-megamoe.yaml | ||
| # | ||
| # Topology: 3 prefill (DEP=8 each) + 1 decode (DEP=8). 9 nodes total with a | ||
| # dedicated NATS/etcd infra node. MegaMOE max-throughput point at concurrency | ||
| # 4096 with no CPU/NVMe offload. | ||
| # | ||
| # Local deltas vs upstream: | ||
| # * model.path alias renamed deepseekv4-fp4 -> deepseek-v4-pro to match | ||
| # SRT_SLURM_MODEL_PREFIX in runners/launch_gb200-nv.sh. | ||
| # * model.container set to vllm/vllm-openai:v0.20.0-ubuntu2404 to | ||
| # match nvidia-master.yaml image (which the launch script registers as | ||
| # the alias key in srtslurm.yaml). Upstream variants ship either the | ||
| # non-dynamo floating tag or a sha256 pin. | ||
| # * slurm.time_limit + health_check set to 8h / 1440 attempts to | ||
| # absorb cold-cache /mnt/numa1 model loads. | ||
| model: | ||
| path: "deepseek-v4-pro" | ||
| container: "vllm/vllm-openai:v0.20.0-ubuntu2404" | ||
| precision: "fp4" | ||
|
|
||
| dynamo: | ||
| install: true | ||
| wheel: "1.2.0.dev20260426" | ||
|
|
||
| setup_script: vllm-container-deps.sh | ||
|
|
||
| slurm: | ||
| time_limit: "8:00:00" | ||
|
|
||
| health_check: | ||
| max_attempts: 1440 | ||
| interval_seconds: 10 | ||
| resources: | ||
| gpu_type: "gb200" | ||
| gpus_per_node: 4 | ||
| prefill_nodes: 6 | ||
| decode_nodes: 2 | ||
| prefill_workers: 3 | ||
| decode_workers: 1 | ||
| gpus_per_prefill: 8 | ||
| gpus_per_decode: 8 | ||
|
|
||
| infra: | ||
| etcd_nats_dedicated_node: true | ||
|
|
||
| frontend: | ||
| type: dynamo | ||
| enable_multiple_frontends: false | ||
| backend: | ||
| type: vllm | ||
| connector: null | ||
| prefill_environment: | ||
| TILELANG_CLEANUP_TEMP_FILES: "1" | ||
| VLLM_USE_NCCL_SYMM_MEM: "1" | ||
| TORCH_SYMMMEM: "NVSHMEM" | ||
| NCCL_CUMEM_ENABLE: "1" | ||
| NCCL_MNNVL_ENABLE: "1" | ||
| NCCL_NVLS_ENABLE: "1" | ||
| VLLM_SERVER_DEV_MODE: "1" | ||
| VLLM_SPARSE_INDEXER_MAX_LOGITS_MB: "1024" | ||
| VLLM_MAX_TOKENS_PER_EXPERT_FP4_MOE: "2048" | ||
| # VLLM_RANDOMIZE_DP_DUMMY_INPUTS: "1" | ||
| # VLLM_MOE_ROUTING_SIMULATION_STRATEGY: "uniform_random" | ||
| UCX_MEMTYPE_CACHE: "n" | ||
| UCX_MEMTYPE_REG_WHOLE: "n" | ||
| UCX_TLS: "cuda_copy,cuda_ipc,tcp" | ||
| UCX_CUDA_IPC_ENABLE_MNNVL: "y" | ||
| NCCL_P2P_LEVEL: NVL | ||
| decode_environment: | ||
| TILELANG_CLEANUP_TEMP_FILES: "1" | ||
| VLLM_USE_NCCL_SYMM_MEM: "1" | ||
| TORCH_SYMMMEM: "NVSHMEM" | ||
| NCCL_CUMEM_ENABLE: "1" | ||
| NCCL_MNNVL_ENABLE: "1" | ||
| NCCL_NVLS_ENABLE: "1" | ||
| VLLM_SERVER_DEV_MODE: "1" | ||
| # VLLM_RANDOMIZE_DP_DUMMY_INPUTS: "1" | ||
| # VLLM_MOE_ROUTING_SIMULATION_STRATEGY: "uniform_random" | ||
| UCX_MEMTYPE_CACHE: "n" | ||
| UCX_MEMTYPE_REG_WHOLE: "n" | ||
| UCX_TLS: "cuda_copy,cuda_ipc,tcp" | ||
| UCX_CUDA_IPC_ENABLE_MNNVL: "y" | ||
| NCCL_P2P_LEVEL: NVL | ||
| vllm_config: | ||
| prefill: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| served-model-name: "deepseek-ai/DeepSeek-V4-Pro" | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 1 | ||
| pipeline-parallel-size: 1 | ||
| data-parallel-size: 8 | ||
| data-parallel-rpc-port: 13345 | ||
| enable-expert-parallel: true | ||
| enable-ep-weight-filter: true | ||
| moe-backend: deep_gemm_mega_moe | ||
| enforce-eager: true | ||
| max-model-len: 9280 | ||
| max-num-seqs: 16 | ||
| max-num-batched-tokens: 32768 | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| no-enable-flashinfer-autotune: true | ||
| no-async-scheduling: true | ||
| block-size: 256 | ||
| gpu-memory-utilization: 0.95 | ||
| no-disable-hybrid-kv-cache-manager: true | ||
| enable-sleep-mode: true | ||
| numa-bind: true | ||
| tokenizer-mode: deepseek_v4 | ||
| decode: | ||
| kv-transfer-config: '{"kv_connector": "NixlConnector", "kv_role": "kv_both"}' | ||
| served-model-name: "deepseek-ai/DeepSeek-V4-Pro" | ||
| kv-cache-dtype: "fp8" | ||
| tensor-parallel-size: 1 | ||
| pipeline-parallel-size: 1 | ||
| data-parallel-size: 8 | ||
| data-parallel-rpc-port: 13345 | ||
| enable-expert-parallel: true | ||
| enable-ep-weight-filter: true | ||
| moe-backend: deep_gemm_mega_moe | ||
| max-model-len: 9280 | ||
|
alec-flowers marked this conversation as resolved.
|
||
| max-num-seqs: 512 | ||
| max-cudagraph-capture-size: 512 | ||
| max-num-batched-tokens: 512 | ||
| trust-remote-code: true | ||
| no-enable-prefix-caching: true | ||
| block-size: 256 | ||
| compilation-config: '{"cudagraph_mode":"FULL_DECODE_ONLY","mode":0}' | ||
| gpu-memory-utilization: 0.9 | ||
| stream-interval: 50 | ||
| no-disable-hybrid-kv-cache-manager: true | ||
| enable-sleep-mode: true | ||
| tokenizer-mode: deepseek_v4 | ||
| benchmark: | ||
| type: "sa-bench" | ||
| isl: 8192 | ||
| osl: 1024 | ||
| concurrencies: "4096" | ||
| req_rate: "inf" | ||
| use_chat_template: true | ||
| custom_tokenizer: "sa_bench_tokenizers.vllm_deepseek_v4.VLLMDeepseekV4Tokenizer" | ||
|
|
||
| identity: | ||
| model: | ||
| repo: "deepseek-ai/DeepSeek-V4-Pro" | ||
| revision: "0366e4e064385807ea86b088a5c6c878ff23343b" | ||
| container: | ||
| image: "vllm/vllm-openai:v0.20.0-ubuntu2404" | ||
| frameworks: | ||
| dynamo: "1.2.0.dev20260426" | ||
| vllm: "0.20.0" | ||
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
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.
Uh oh!
There was an error while loading. Please reload this page.