diff --git a/docs/source/tutorials/multi-node_dsv3.2.md b/docs/source/tutorials/multi-node_dsv3.2.md index 0cd7ef07f..39633e441 100644 --- a/docs/source/tutorials/multi-node_dsv3.2.md +++ b/docs/source/tutorials/multi-node_dsv3.2.md @@ -192,7 +192,7 @@ Before launch the inference server, ensure the following environment variables a #!/bin/sh # this obtained through ifconfig -# nic_name is the network interface name corresponding to local_ip +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" @@ -230,9 +230,14 @@ vllm serve /root/.cache/Modelers_Park/DeepSeek-V3.2-Exp \ ```shell #!/bin/sh +# this obtained through ifconfig +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxx" local_ip="xxx" +# The value of node0_ip must be consistent with the value of local_ip set in node0 (master node) +node0_ip="xxxx" + export VLLM_USE_MODELSCOPE=True export HCCL_IF_IP=$local_ip export GLOO_SOCKET_IFNAME=$nic_name @@ -249,7 +254,7 @@ vllm serve /root/.cache/Modelers_Park/DeepSeek-V3.2-Exp \ --data-parallel-size 2 \ --data-parallel-size-local 1 \ --data-parallel-start-rank 1 \ ---data-parallel-address \ +--data-parallel-address $node0_ip \ --data-parallel-rpc-port 13389 \ --tensor-parallel-size 16 \ --seed 1024 \ @@ -300,7 +305,7 @@ Run the following scripts on two nodes respectively #!/bin/sh # this obtained through ifconfig -# nic_name is the network interface name corresponding to local_ip +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" @@ -341,9 +346,14 @@ vllm serve vllm-ascend/DeepSeek-V3.2-Exp-W8A8 \ ```shell #!/bin/sh +# this obtained through ifconfig +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxx" local_ip="xxx" +# The value of node0_ip must be consistent with the value of local_ip set in node0 (master node) +node0_ip="xxxx" + export VLLM_USE_MODELSCOPE=True export HCCL_IF_IP=$local_ip export GLOO_SOCKET_IFNAME=$nic_name @@ -362,7 +372,7 @@ vllm serve vllm-ascend/DeepSeek-V3.2-Exp-W8A8 \ --data-parallel-size 2 \ --data-parallel-size-local 1 \ --data-parallel-start-rank 1 \ ---data-parallel-address \ +--data-parallel-address $node0_ip \ --data-parallel-rpc-port 13389 \ --tensor-parallel-size 8 \ --seed 1024 \ diff --git a/docs/source/tutorials/multi_node.md b/docs/source/tutorials/multi_node.md index e4ca3c59f..00a61a67e 100644 --- a/docs/source/tutorials/multi_node.md +++ b/docs/source/tutorials/multi_node.md @@ -103,7 +103,7 @@ Before launch the inference server, ensure the following environment variables a #!/bin/sh # this obtained through ifconfig -# nic_name is the network interface name corresponding to local_ip +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" @@ -144,9 +144,14 @@ vllm serve vllm-ascend/DeepSeek-V3.1-W8A8 \ ```shell #!/bin/sh +# this obtained through ifconfig +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxx" local_ip="xxx" +# The value of node0_ip must be consistent with the value of local_ip set in node0 (master node) +node0_ip="xxxx" + export VLLM_USE_MODELSCOPE=True export HCCL_IF_IP=$local_ip export GLOO_SOCKET_IFNAME=$nic_name @@ -164,7 +169,7 @@ vllm serve vllm-ascend/DeepSeek-V3.1-W8A8 \ --data-parallel-size 4 \ --data-parallel-size-local 2 \ --data-parallel-start-rank 2 \ ---data-parallel-address { node0 ip } \ +--data-parallel-address $node0_ip \ --data-parallel-rpc-port 13389 \ --tensor-parallel-size 4 \ --seed 1024 \ diff --git a/docs/source/tutorials/multi_node_kimi.md b/docs/source/tutorials/multi_node_kimi.md index dfada8581..09ee2ca25 100644 --- a/docs/source/tutorials/multi_node_kimi.md +++ b/docs/source/tutorials/multi_node_kimi.md @@ -59,7 +59,7 @@ Before launch the inference server, ensure the following environment variables a #!/bin/sh # this obtained through ifconfig -# nic_name is the network interface name corresponding to local_ip +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" @@ -101,9 +101,14 @@ vllm serve /home/cache/weights/Kimi-K2-Instruct-W8A8 \ ```shell #!/bin/sh +# this obtained through ifconfig +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" +# The value of node0_ip must be consistent with the value of local_ip set in node0 (master node) +node0_ip="xxxx" + export HCCL_IF_IP=$local_ip export GLOO_SOCKET_IFNAME=$nic_name export TP_SOCKET_IFNAME=$nic_name diff --git a/docs/source/tutorials/multi_node_qwen3vl.md b/docs/source/tutorials/multi_node_qwen3vl.md index a9fe3b8e3..18d063d4a 100644 --- a/docs/source/tutorials/multi_node_qwen3vl.md +++ b/docs/source/tutorials/multi_node_qwen3vl.md @@ -59,7 +59,7 @@ node0 ```shell #!/bin/sh # this obtained through ifconfig -# nic_name is the network interface name corresponding to local_ip +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" @@ -97,8 +97,12 @@ node1 ```shell #!/bin/sh +# this obtained through ifconfig +# nic_name is the network interface name corresponding to local_ip of the current node nic_name="xxxx" local_ip="xxxx" + +# The value of node0_ip must be consistent with the value of local_ip set in node0 (master node) node0_ip="xxxx" export HCCL_IF_IP=$local_ip