最新版的2.1.RC1-800I-A2-py311-openeuler24.03-lts 部署Qwen3-235B-A22B （2台Atlas 800I A2 推理）目前推理速度10 token/s_MindIE_昇腾论坛

# 最新版的2.1.RC1-800I-A2-py311-openeuler24.03-lts 部署Qwen3-235B-A22B （2台Atlas 800I A2 推理）目前推理速度10 token/s_MindIE_昇腾论坛

## 概述

本文档基于昇腾社区论坛帖子生成的技术教程。

**原始链接**: https://www.hiascend.com/forum/thread-0259189517084820024-1-1.html

**生成时间**: 2025-08-27 10:33:57

---

## 问题描述

单次请求的推理速度10 token/s能否提高速度怎么提高有没有大佬指点一下 镜像版本config.json和环境变量可以发一下 发到2楼了 镜像2.1.RC1-800I-A2-py311-openeuler24.03-lts export ATB_LLM_HCCL_ENABLE=1 export ATB_LLM_COMM_BACKEND="hccl" export HCCL_CONNECT_TIMEOUT=7200 # 该环境变量需要配置为整数取值范围[120,7200]单位s export HCCL_EXEC_TIMEOUT=0 export PYTORCH_NPU_ALLOC_CONF=expandable_segments:True export ASCEND_RT_VISIBLE_DEVICES="0,1,2,3,4,5,6,7" { "Version" : "1.0.0", "ServerConfig" : { "ipAddress" : "192.168.1.3", "managementIpAddress" : "192.168.1.3", "port" : 1025, "managementPort" : 1026, "metricsPort" : 1027, "allowAllZeroIpListening" : false, "maxLinkNum" : 50, "httpsEnabled" : false, "fullTextEnabled" : false, "tlsCaPath" : "security/ca/", "tlsCaFile" : ["ca.pem"], "tlsCert" : "security/certs/server.pem", "tlsPk" : "security/keys/server.key.pem", "tlsPkPwd" : "security/pass/key_pwd.txt", "tlsCrlPath" : "security/certs/", "tlsCrlFiles" : ["server_crl.pem"], "managementTlsCaFile" : ["management_ca.pem"], "managementTlsCert" : "security/certs/management/...

## 完整内容

单次请求的推理速度10 token/s能否提高速度怎么提高有没有大佬指点一下 镜像版本config.json和环境变量可以发一下 发到2楼了 镜像2.1.RC1-800I-A2-py311-openeuler24.03-lts export ATB_LLM_HCCL_ENABLE=1 export ATB_LLM_COMM_BACKEND="hccl" export HCCL_CONNECT_TIMEOUT=7200 # 该环境变量需要配置为整数取值范围[120,7200]单位s export HCCL_EXEC_TIMEOUT=0 export PYTORCH_NPU_ALLOC_CONF=expandable_segments:True export ASCEND_RT_VISIBLE_DEVICES="0,1,2,3,4,5,6,7" { "Version" : "1.0.0", "ServerConfig" : { "ipAddress" : "192.168.1.3", "managementIpAddress" : "192.168.1.3", "port" : 1025, "managementPort" : 1026, "metricsPort" : 1027, "allowAllZeroIpListening" : false, "maxLinkNum" : 50, "httpsEnabled" : false, "fullTextEnabled" : false, "tlsCaPath" : "security/ca/", "tlsCaFile" : ["ca.pem"], "tlsCert" : "security/certs/server.pem", "tlsPk" : "security/keys/server.key.pem", "tlsPkPwd" : "security/pass/key_pwd.txt", "tlsCrlPath" : "security/certs/", "tlsCrlFiles" : ["server_crl.pem"], "managementTlsCaFile" : ["management_ca.pem"], "managementTlsCert" : "security/certs/management/server.pem", "managementTlsPk" : "security/keys/management/server.key.pem", "managementTlsPkPwd" : "security/pass/management/key_pwd.txt", "managementTlsCrlPath" : "security/management/certs/", "managementTlsCrlFiles" : ["server_crl.pem"], "kmcKsfMaster" : "tools/pmt/master/ksfa", "kmcKsfStandby" : "tools/pmt/standby/ksfb", "inferMode" : "standard", "interCommTLSEnabled" : false, "interCommPort" : 1121, "interCommTlsCaPath" : "security/grpc/ca/", "interCommTlsCaFiles" : ["ca.pem"], "interCommTlsCert" : "security/grpc/certs/server.pem", "interCommPk" : "security/grpc/keys/server.key.pem", "interCommPkPwd" : "security/grpc/pass/key_pwd.txt", "interCommTlsCrlPath" : "security/grpc/certs/", "interCommTlsCrlFiles" : ["server_crl.pem"], "openAiSupport" : "vllm", "tokenTimeout" : 600, "e2eTimeout" : 600, "distDPServerEnabled":false }, "BackendConfig" : { "backendName" : "mindieservice_llm_engine", "modelInstanceNumber" : 1, "npuDeviceIds" : [[0,1,2,3,4,5,6,7]], "tokenizerProcessNumber" : 8, "multiNodesInferEnabled" : true, "multiNodesInferPort" : 1120, "interNodeTLSEnabled" : false, "interNodeTlsCaPath" : "security/grpc/ca/", "interNodeTlsCaFiles" : ["ca.pem"], "interNodeTlsCert" : "security/grpc/certs/server.pem", "interNodeTlsPk" : "security/grpc/keys/server.key.pem", "interNodeTlsPkPwd" : "security/grpc/pass/mindie_server_key_pwd.txt", "interNodeTlsCrlPath" : "security/grpc/certs/", "interNodeTlsCrlFiles" : ["server_crl.pem"], "interNodeKmcKsfMaster" : "tools/pmt/master/ksfa", "interNodeKmcKsfStandby" : "tools/pmt/standby/ksfb", "ModelDeployConfig" : { "maxSeqLen" : 42768, "maxInputTokenLen" : 24566, "truncation" : true, "ModelConfig" : [ { "modelInstanceType" : "Standard", "modelName" : "qwen3_235b", "modelWeightPath" : "/data/Qwen3-235B-A22B", "worldSize" : 8, "cpuMemSize" : 5, "npuMemSize" : -1, "backendType" : "atb", "trustRemoteCode" : false } ] }, "ScheduleConfig" : { "templateType" : "Standard", "templateName" : "Standard_LLM", "cacheBlockSize" : 128, "maxPrefillBatchSize" : 256, "maxPrefillTokens" : 48768, "prefillTimeMsPerReq" : 600, "prefillPolicyType" : 0, "decodeTimeMsPerReq" : 50, "decodePolicyType" : 0, "maxBatchSize" : 512, "maxIterTimes" : 18192, "maxPreemptCount" : 0, "supportSelectBatch" : false, "maxQueueDelayMicroseconds" : 5000 } } }

---

## 技术要点总结

基于以上内容，主要技术要点包括：

1. **问题类型**: 部署安装
2. **涉及技术**: HTTPS, PyTorch, NPU, Atlas, MindIE, 昇腾, AI
3. **解决方案**: 请参考完整内容中的解决方案

## 相关资源

- 昇腾社区: https://www.hiascend.com/
- 昇腾论坛: https://www.hiascend.com/forum/

---

*本文档由AI自动生成，仅供参考。如有疑问，请参考原始帖子。*