Azure 机器学习服务限制Service limits in Azure Machine Learning
本部分列出了 Azure 机器学习中的基本配额和限制阈值。This section lists basic quotas and throttling thresholds in Azure Machine Learning. 若要了解如何增加资源配额,请参阅管理和增加资源配额To learn how increase resource quotas, see "Manage and increase quotas for resources"
工作区Workspaces
限制Limit | ValueValue |
---|---|
工作区名称Workspace name | 2-32 个字符2-32 characters |
运行次数Runs
限制Limit | ValueValue |
---|---|
每工作区运行数Runs per workspace | 1 千万10 million |
RunId/ParentRunIdRunId/ParentRunId | 256 个字符256 characters |
DataContainerIdDataContainerId | 261 个字符261 characters |
DisplayNameDisplayName | 256 个字符256 characters |
说明Description | 5,000 个字符5,000 characters |
属性数Number of properties | 5050 |
属性键的长度Length of property key | 100 个字符100 characters |
属性值的长度Length of property value | 1,000 个字符1,000 characters |
标记数Number of tags | 5050 |
标记键长度Length of tag key | 100100 |
标记值长度Length of tag value | 1,000 个字符1,000 characters |
CancelUri / CompleteUri / DiagnosticsUriCancelUri / CompleteUri / DiagnosticsUri | 1,000 个字符1,000 characters |
错误消息长度Error message length | 3,000 个字符3,000 characters |
警告消息长度Warning message length | 300 个字符300 characters |
输入数据集数Number of input datasets | 200200 |
输出数据集数Number of output datasets | 2020 |
指标Metrics
限制Limit | ValueValue |
---|---|
每次运行的指标名称Metric names per run | 5050 |
每个指标的指标行名称Metric rows per metric name | 1 千万10 million |
每个指标行的列数Columns per metric row | 1515 |
指标列名称长度Metric column name length | 255 个字符255 characters |
指标列值长度Metric column value length | 255 个字符255 characters |
每批上传的指标行数Metric rows per batch uploaded | 250250 |
备注
如果由于将变量格式化为指标名称而达到每个运行的指标名称限制,请考虑改用行指标,其中一列是变量值,第二列是指标值。If you are hitting the limit of metric names per run because you are formatting variables into the metric name, consider instead to use a row metric where one column is the variable value and the second column is the metric value.
ArtifactsArtifacts
限制Limit | ValueValue |
---|---|
每次运行的项目数Number of artifacts per run | 1 千万10 million |
项目路径的最大长度Max length of artifact path | 5,000 个字符5,000 characters |
限制增加Limit increases
可以通过联系支持人员来提升单个工作区的一些限制。Some limits can be increased for individual workspaces by contacting support.
后续步骤Next steps
- 配置 Azure 机器学习环境Configure your Azure Machine Learning environment
- 若要了解如何增加资源配额,请参阅管理和增加资源配额。Learn how increase resource quotas in "Manage and increase quotas for resources".