创建训练任务模版

将一组常用的训练配置（资源规格、镜像、存储挂载、环境变量、启动命令等）保存为模版，后续创建训练任务时可直接复用，避免重复填写。创建成功后可在训练任务模版列表中查看，也可在创建分布式训练任务时以 createType=TEMPLATE 引用。

POSThttps://api.alayanew.com/v1/training/template/create

鉴权（Authorizations）

bearerAuth

AuthorizationString必填

用户可通过已获取的 Open API Key 做验证。例如：Bearer [YOUR_API_KEY]。

Request body

application/json

nameString

模版名称，用于在模版列表中区分识别。例如：pytorch-2gpu-template。

descString

模版描述，补充说明该模版的适用场景。例如：常用2卡训练模板。

aidcIdInteger

智算中心 ID，指定模版默认绑定的智算中心。例如：2。

trainingTypeString

训练类型，可选值：PRE_TRAINING 预训练、HPC 高性能计算。例如：PRE_TRAINING。

trainingFramworkString

训练框架，可选值：PyTorch、DeepSpeed、MPI、TensorFlow。例如：TensorFlow。

imageTypeString

镜像类型，可选值：general 基础镜像、application 应用镜像、private 私有镜像。需与 image 字段对应。例如：general。

imageString

容器镜像地址（用户选择的容器镜像）。例如：harbor.zetyun.cn/anc-public/general/tensorflow:2.16.1-gpu-jupyter。

envObject

环境变量（键值对）。

显示 properties

emptyBoolean

Additional propertiesObject

额外参数，例如："ENV_MODE":"production"，"MAX_CONN":"200"。

resourceObject

资源配置。

显示 properties

emptyBoolean

Additional propertiesObject

额外参数，例如："type":"worker"，"gpuName":"NVIDIA-L40S-PCIE-48G"，"cpuCores":"1"，"gpuCount":"1"，"memoryGB":"2"，"productCode":"PRD-QTT"，"productName":"量子训练"，"workerCount":1，"productPrice":"0.0002B"。

enableAutoRetryBoolean

是否支持自动重试。

storageConfigsArray<Object>

存储配置参数。

显示 properties

emptyBoolean

ItemsObject

例如："storageId":"0000-0000-0000-0000"，"storageType":"nas-capacity"，"fileDirectory":"nas123"，"mountPath":"/root/nas/123"，"onlyRead":true。

maxRetryCountInteger

最大重试次数，例如：3。

enableTimeoutCancelBoolean

是否支持超时取消。例如：true。

timeoutHoursInteger

训练任务超时时间（单位：小时），例如：3。

startCommandString

启动命令，例如："python train.py --data /root/nas/123/data"。

priorityInteger

优先级，取值范围：1-3，例如：3。

Response

application/json · 200

statusInteger

业务状态码，200 表示成功。

messageString

接口响应信息。例如："OK"。

dataObject

创建结果。创建成功时不返回额外业务数据。

curl -X 'POST' \
  'https://api.alayanew.com/v1/training/template/create' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer [YOUR_API_KEY]' \
  -d '{
    "name": "my training task",
    "desc": "my training task",
    "aidcId": 2,
    "trainingType": "PRE_TRAINING",
    "trainingFramwork": "TensorFlow",
    "imageType": "general",
    "image": "harbor.zetyun.cn/anc-public/general/tensorflow:2.16.1-gpu-jupyter",
    "env": {
      "ENV_MODE": "production",
      "MAX_CONN": "200"
    },
    "resource": {
      "type": "worker",
      "gpuName": "NVIDIA-L40S-PCIE-48G",
      "cpuCores": "1",
      "gpuCount": "1",
      "memoryGB": "2",
      "productCode": "PRD-QTT",
      "productName": "量子训练",
      "workerCount": 1,
      "productPrice": "0.0002B"
    },
    "enableAutoRetry": true,
    "storageConfigs": [
      {
        "storageId": "0000-0000-0000-0000",
        "storageType": "nas-capacity",
        "fileDirectory": "nas123",
        "mountPath": "/root/nas/123",
        "onlyRead": true
      }
    ],
    "maxRetryCount": 3,
    "enableTimeoutCancel": true,
    "timeoutHours": 1,
    "startCommand": "python train.py --data /root/nas/123/data",
    "priority": 3
  }'

import requests

url = "https://api.alayanew.com/v1/training/template/create"
headers = {
    "accept": "application/json",
    "Content-Type": "application/json",
    "Authorization": "Bearer [YOUR_API_KEY]"
}
payload = {
    "name": "pytorch-2gpu-template",
    "desc": "常用2卡训练模板",
    "aidcId": 2,
    "trainingType": "PRE_TRAINING",
    "trainingFramwork": "TensorFlow",
    "imageType": "general",
    "image": "harbor.zetyun.cn/anc-public/general/tensorflow:2.16.1-gpu-jupyter",
    "env": {"ENV_MODE": "production", "MAX_CONN": "200"},
    "resource": {
        "type": "worker",
        "gpuName": "NVIDIA-L40S-PCIE-48G",
        "cpuCores": "1",
        "gpuCount": "1",
        "memoryGB": "2",
        "productCode": "PRD-QTT",
        "productName": "量子训练",
        "workerCount": 1,
        "productPrice": "0.0002B"
    },
    "storageConfigs": [
        {
            "storageId": "0000-0000-0000-0000",
            "storageType": "nas-capacity",
            "fileDirectory": "nas123",
            "mountPath": "/root/nas/123",
            "onlyRead": True
        }
    ],
    "enableAutoRetry": True,
    "maxRetryCount": 3,
    "enableTimeoutCancel": True,
    "timeoutHours": 1,
    "startCommand": "python train.py --data /root/nas/123/data",
    "priority": 3
}

response = requests.post(url, headers=headers, json=payload)
response.raise_for_status()
print(response.json())

const payload = {
  name: 'pytorch-2gpu-template',
  desc: '常用2卡训练模板',
  aidcId: 2,
  trainingType: 'PRE_TRAINING',
  trainingFramwork: 'TensorFlow',
  imageType: 'general',
  image: 'harbor.zetyun.cn/anc-public/general/tensorflow:2.16.1-gpu-jupyter',
  env: { ENV_MODE: 'production', MAX_CONN: '200' },
  resource: {
    type: 'worker',
    gpuName: 'NVIDIA-L40S-PCIE-48G',
    cpuCores: '1',
    gpuCount: '1',
    memoryGB: '2',
    productCode: 'PRD-QTT',
    productName: '量子训练',
    workerCount: 1,
    productPrice: '0.0002B'
  },
  storageConfigs: [
    {
      storageId: '0000-0000-0000-0000',
      storageType: 'nas-capacity',
      fileDirectory: 'nas123',
      mountPath: '/root/nas/123',
      onlyRead: true
    }
  ],
  enableAutoRetry: true,
  maxRetryCount: 3,
  enableTimeoutCancel: true,
  timeoutHours: 1,
  startCommand: 'python train.py --data /root/nas/123/data',
  priority: 3
};

fetch('https://api.alayanew.com/v1/training/template/create', {
  method: 'POST',
  headers: {
    'accept': 'application/json',
    'Content-Type': 'application/json',
    'Authorization': 'Bearer [YOUR_API_KEY]'
  },
  body: JSON.stringify(payload)
})
  .then(res => {
    if (!res.ok) {
      throw new Error(`HTTP error! status: ${res.status}`);
    }
    return res.json();
  })
  .then(console.log)
  .catch(console.error);

{
  "status": 200,
  "message": "OK",
  "data": {}
}

{
  "status": 403,
  "message": "Forbidden",
  "data": {}
}

{
  "status": 500,
  "message": "Internal Server Error",
  "data": {}
}