一、环境准备二、工作区限制三、什么是计算目标四、本地计算机五、远程虚拟机六、Apache

Spark

关注TechLead分享AI全维度知识。

作者拥有10年互联网服务架构、AI产品研发经验、团队管理经验同济本复旦硕复旦机器人智能实验室成员阿里云认证的资深架构师项目管理专业人士上亿营收AI产品研发负责人。

一、环境准备

LocalWebservice.deploy_configuration()

来定义

ID/subscriptions/subscription_id/resourceGroups/resource_group/providers/Microsoft.Compute/virtualMachines/vm_name

from

RemoteCompute.attach_configuration(resource_idresource_id,ssh_port22,usernameusername,passwordpassword)#

Attach

attach_config)compute.wait_for_completion(show_outputTrue)或者可以使用

Azure

azureml.core.conda_dependencies

import

myenv.python.conda_dependencies

CondaDependencies.create(conda_packages[scikit-learn])#

base

azureml.core.runconfig.DEFAULT_CPU_IMAGE

will

azureml.core.runconfig.DEFAULT_GPU_IMAGE#

Configure

ScriptRunConfig(source_directory.,

scripttrain.py,

clustername-ssh.azurehdinsight.net

附加若要将

ID/subscriptions/subscription_id/resourceGroups/resource_group/providers/Microsoft.HDInsight/clusters/cluster_name

from

private_key_passphraseattach_config

HDInsightCompute.attach_configuration(resource_idresource_id,ssh_port22,

ComputeTarget.attach(workspacews,

namemyhdi,

attach_configurationattach_config)except

e:print(Caught

{}.format(e.message))hdi_compute.wait_for_completion(show_outputTrue)或者可以使用

Azure

azureml.core.conda_dependencies

import

RunConfiguration(frameworkpyspark)#

Set

run_hdi.environment.python.conda_dependencies

群集请使用

ComputeTargetException:print(Attaching

Batch

BatchCompute.attach_configuration(resource_groupbatch_resource_group,

account_namebatch_account_name)batch_compute

batch_compute_name,

provisioning_config)batch_compute.wait_for_completion()print(Provisioning

state:{}.format(batch_compute.provisioning_state))print(Provisioning

errors:{}.format(batch_compute.provisioning_errors))print(Using

Batch

compute:{}.format(batch_compute.cluster_resource_id))请勿在工作区中为同一

Azure

ComputeTargetExceptiondatabricks_compute_name

os.environ.get(AML_DATABRICKS_COMPUTE_NAME,

os.environ.get(AML_DATABRICKS_WORKSPACE,

os.environ.get(AML_DATABRICKS_RESOURCE_GROUP,

os.environ.get(AML_DATABRICKS_ACCESS_TOKEN,

databricks_access_token)try:databricks_compute

namedatabricks_compute_name)print(Compute

target

ComputeTargetException:print(compute

not

found)print(databricks_compute_name

{}.format(databricks_compute_name))print(databricks_workspace_name

{}.format(databricks_workspace_name))print(databricks_access_token

{}.format(databricks_access_token))#

Create

DatabricksCompute.attach_configuration(resource_groupdatabricks_resource_group,workspace_namedatabricks_workspace_name,access_tokendatabricks_access_token)databricks_compute

ComputeTarget.attach(ws,databricks_compute_name,attach_config)databricks_compute.wait_for_completion(True)有关更详细的示例请参阅

GitHub

ComputeTargetExceptionadla_compute_name

os.environ.get(AML_ADLA_COMPUTE_NAME,

adla_compute_name)

os.environ.get(AML_ADLA_RESOURCE_GROUP,

adla_resource_group)

os.environ.get(AML_ADLA_ACCOUNT_NAME,

adla_account_name)try:adla_compute

nameadla_compute_name)print(Compute

target

ComputeTargetException:print(compute

not

{}.format(adla_compute_name))print(adla_resource_id

{}.format(adla_resource_group))print(adla_account_name

create

AdlaCompute.attach_configuration(resource_groupadla_resource_group,account_nameadla_account_name)#

Attach

ComputeTarget.attach(ws,adla_compute_name,attach_config)adla_compute.wait_for_completion(True)有关更详细的示例请参阅

GitHub

compute_target.detach()关注TechLead分享AI全维度知识。

作者拥有10年互联网服务架构、AI产品研发经验、团队管理经验同济本复旦硕复旦机器人智能实验室成员阿里云认证的资深架构师项目管理专业人士上亿营收AI产品研发负责人。