835b4c8f1d
aclgraph is stable and fast now. Let's drop torchair graph mode now.
TODO: some logic to adapt torchair should be cleaned up as well. We'll
do it in the following PR.
- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Co-authored-by: Mengqing Cao <cmq0113@163.com>
52 lines
1.4 KiB
YAML
52 lines
1.4 KiB
YAML
test_name: "test DeepSeek-V3.2-Exp-bf16 multi-dp"
|
|
model: "Yanguan/DeepSeek-V3.2-Exp-bf16"
|
|
num_nodes: 2
|
|
npu_per_node: 16
|
|
env_common:
|
|
VLLM_USE_MODELSCOPE: true
|
|
OMP_PROC_BIND: false
|
|
OMP_NUM_THREADS: 100
|
|
HCCL_BUFFSIZE: 1024
|
|
SERVER_PORT: 8080
|
|
VLLM_ASCEND_ENABLE_MLAPO: 0
|
|
|
|
deployment:
|
|
-
|
|
server_cmd: >
|
|
vllm serve "Yanguan/DeepSeek-V3.2-Exp-bf16"
|
|
--host 0.0.0.0
|
|
--port $SERVER_PORT
|
|
--data-parallel-address $LOCAL_IP
|
|
--data-parallel-size 2
|
|
--data-parallel-size-local 1
|
|
--data-parallel-rpc-port 13389
|
|
--tensor-parallel-size 16
|
|
--seed 1024
|
|
--enable-expert-parallel
|
|
--max-num-seqs 16
|
|
--max-model-len 17450
|
|
--max-num-batched-tokens 17450
|
|
--trust-remote-code
|
|
--no-enable-prefix-caching
|
|
--gpu-memory-utilization 0.9
|
|
|
|
-
|
|
server_cmd: >
|
|
vllm serve "Yanguan/DeepSeek-V3.2-Exp-bf16"
|
|
--headless
|
|
--data-parallel-size 2
|
|
--data-parallel-size-local 1
|
|
--data-parallel-start-rank 1
|
|
--data-parallel-address $MASTER_IP
|
|
--data-parallel-rpc-port 13389
|
|
--tensor-parallel-size 16
|
|
--seed 1024
|
|
--max-num-seqs 16
|
|
--max-model-len 17450
|
|
--max-num-batched-tokens 17450
|
|
--enable-expert-parallel
|
|
--trust-remote-code
|
|
--no-enable-prefix-caching
|
|
--gpu-memory-utilization 0.92
|
|
benchmarks:
|