示例代码

在ModelArts notebook平台，Session鉴权无需输入鉴权参数。其它平台的Session鉴权请参见Session鉴权。

• 方式1：已部署为在线服务predictor的初始化

  1 2 3 4 5

  from modelarts.session import Session from modelarts.model import Predictor session = Session() predictor_instance = Predictor(session, service_id="your_service_id")

• 方式2：部署在线服务predictor
  • 部署服务到公共资源池

    1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

    from modelarts.session import Session from modelarts.model import Model from modelarts.config.model_config import ServiceConfig, TransformerConfig, Schedule session = Session() model_instance = Model(session, model_id='your_model_id') vpc_id = None # （可选）在线服务实例部署的虚拟私有云ID，默认为空 subnet_network_id = None # （可选）子网的网络ID，默认为空 security_group_id = None # （可选）安全组，默认为空 configs = [ServiceConfig(model_id=model_instance.model_id, weight="100", instance_count=1, specification="modelarts.vm.cpu.2u")] # 参考表3中specification字段 predictor_instance = model_instance.deploy_predictor( service_name="service_predictor_name", infer_type="real-time", vpc_id=vpc_id, subnet_network_id=subnet_network_id, security_group_id=security_group_id, configs=configs, # predictor配置参数, 参考下文configs参数格式说明 schedule = [Schedule(op_type='stop', time_unit='HOURS', duration=1)] # （可选）设置在线服务运行时间 )

    参数“model_id”代表将部署成在线服务的模型。“model_id”可以通过查询模型列表或者ModelArts管理控制台获取。

  • 部署服务到专属资源池

    from modelarts.config.model_config import ServiceConfig
    
    configs = [ServiceConfig(model_id=model_instance.model_id, weight="100", instance_count=1, 
    						 specification="modelarts.vm.cpu.2u")]
    predictor_instance = model_instance.deploy_predictor( 
                                                          service_name="your_service_name",
                                                          infer_type="real-time",
                                                          configs=configs,
                                                          cluster_id="your dedicated pool id"
                                                        )

  configs参数格式说明：SDK提供了ServiceConfig类对其定义，configs为list，list中的元组对象是ServiceConfig。定义代码如下：

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

  configs = [] envs = {"model_name":"mxnet-model-1", "load_epoch":"0"} service_config1 = ServiceConfig( model_id="model_id1", # model_id1和model_id2必须是同一个模型的不同版本对应的model_id weight="70", specification="modelarts.vm.cpu.2u", # 参考表3中specification字段 instance_count=2, envs=envs) # （可选）设置环境变量的值，如：envs = {"model_name":"mxnet-model-1", "load_epoch":"0"} service_config2 = ServiceConfig( model_id='model_id2', weight="30", specification="modelarts.vm.cpu.2u", # 参考表3中specification字段 instance_count=2, envs=envs) # （可选）设置环境变量的值，如：envs = {"model_name":"mxnet-model-1", "load_epoch":"0"} configs.append(service_config1) configs.append(service_config2)

• 方式3：部署批量服务transformer

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

  from modelarts.session import Session from modelarts.model import Model from modelarts.config.model_config import TransformerConfig session = Session() model_instance = Model(session, model_id='your_model_id') vpc_id = None # （可选）批量服务实例部署的虚拟私有云ID，默认为空 subnet_network_id = None # （可选）子网的网络ID，默认为空 security_group_id = None # （可选）安全组，默认为空 transformer = model_instance.deploy_transformer( service_name="service_transformer_name", infer_type="batch", vpc_id=vpc_id, subnet_network_id=subnet_network_id, security_group_id=security_group_id, configs=configs # transformer配置参数, 参考下文configs参数格式说明 )

  configs参数格式说明：SDK提供了TransformerConfig类对其定义，configs都是list，list中的元组对象是TransformerConfig。定义代码如下：

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

  configs = [] mapping_rule = None # （可选）输入参数与csv数据的映射关系 mapping_type= "file" # file或者csv envs = {"model_name":"mxnet-model-1", "load_epoch":"0"} transformer_config1 = TransformerConfig( model_id="model_id", specification="modelarts.vm.cpu.2u", # 参考表3中specification字段 instance_count=2, src_path="/shp-cn4/sdk-demo/", # 批量任务输入数据的OBS路径，如："/your_obs_bucket/src_path" dest_path="/shp-cn4/data-out/", # 批量任务输出结果的OBS路径，如："/your_obs_bucket/dest_path" req_uri="/", mapping_type=mapping_type, mapping_rule=mapping_rule, envs=envs) # （可选）设置环境变量的值，如：envs = {"model_name":"mxnet-model-1", "load_epoch":"0"} configs.append(transformer_config1)