Hand-Qwen3-235B-A22B-W8A8

废弃 本版存在较高频率出现 "游戏副本"字段 、"方块乱码" 已废弃，新的仓库已经修复，请使用 -> jaywoods/Qwen3-235B-A22B-W8A8

您可以通过如下git clone命令，或者ModelScope SDK来下载模型

SDK下载

#安装ModelScope
pip install modelscope

#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('jaywoods/Hand-Qwen3-235B-A22B-W8A8')

Git下载

#Git模型下载
git clone https://www.modelscope.cn/jaywoods/Hand-Qwen3-235B-A22B-W8A8.git

服务化推理

创建mindie容器

对应的镜像介质，从昇腾官方申请获取

docker run -it -d --net=host --shm-size=512g \
    --privileged \
    --name qwen3 \
    -v /usr/local/Ascend/driver:/usr/local/Ascend/driver:ro \
    -v /usr/local/sbin:/usr/local/sbin:ro \
    -v /data/models:/data/models \
    mindie:xxxx bash

配置文件参考

{
    "Version" : "1.0.0",

    "ServerConfig" :
    {
        "ipAddress" : "0.0.0.0",
        "managementIpAddress" : "0.0.0.0",
        "port" : 1025,
        "managementPort" : 1026,
        "metricsPort" : 1027,
        "allowAllZeroIpListening" : true,
        "maxLinkNum" : 1000,
        "httpsEnabled" : false,
        ......
    },

    "BackendConfig" : {
        "backendName" : "mindieservice_llm_engine",
        "modelInstanceNumber" : 1,
        "npuDeviceIds" : [[0,1,2,3,4,5,6,7]],
        "tokenizerProcessNumber" : 8,
        "multiNodesInferEnabled" : false,
        "multiNodesInferPort" : 1120,
        ......
        "ModelDeployConfig" :
        {
            "maxSeqLen" : 16384,
            "maxInputTokenLen" : 8192,
            "truncation" : false,
            "ModelConfig" : [
                {
                    "modelInstanceType" : "Standard",
                    "modelName" : "qwen3-235b",
                    "modelWeightPath" : "/data/jaywoods/Hand-Qwen3-235B-A22B-W8A8",
                    "worldSize" : 8,
                    "cpuMemSize" : 100,
                    "npuMemSize" : -1,
                    "backendType" : "atb",
                    "trustRemoteCode" : true
                }
            ]
        },

        "ScheduleConfig" :
        {
            "templateType" : "Standard",
            "templateName" : "Standard_LLM",
            "cacheBlockSize" : 128,

            "maxPrefillBatchSize" : 50,
            "maxPrefillTokens" : 10000,
            "prefillTimeMsPerReq" : 150,
            "prefillPolicyType" : 0,

            "decodeTimeMsPerReq" : 50,
            "decodePolicyType" : 0,

            "maxBatchSize" : 50,
            "maxIterTimes" : 8192,
            "maxPreemptCount" : 0,
            "supportSelectBatch" : false,
            "maxQueueDelayMicroseconds" : 5000
        }
    }
}

启动

docker exec -it qwen3 bash
cd /usr/local/Ascend/mindie/latest/mindie-service
nohup bin/mindieservice_daemon &

对话测试

curl 'http://127.0.0.1:1025/v1/chat/completions'  \
  -H 'Content-Type: application/json'  -d '{
  "model": "qwen3-235b",
  "stream": false,
  "messages": [
    {
      "role": "user",
      "content": "你是谁?/no_think"
    }
  ]
}'

如果您是本模型的贡献者，我们邀请您根据模型贡献文档，及时完善模型卡片内容。