快速入门 Quick start

MLflow 是用于管理端到端机器学习生命周期的开源平台。MLflow is an open source platform for managing the end-to-end machine learning lifecycle. 它包括三个主要组件:跟踪、模型和项目。It has three primary components: Tracking, Models, and Projects. 通过 MLflow 的“跟踪”组件可使用 JavaPythonRREST API 来记录和查询机器模型训练会话(也称为运行)。The MLflow Tracking component lets you log and query machine model training sessions (runs) using Java, Python, R, and REST APIs. MLflow 运行是与机器学习模型训练过程相关的参数、指标、标记和项目的集合。An MLflow run is a collection of parameters, metrics, tags, and artifacts associated with a machine learning model training process.

试验是 MLflow 中组织的主要单位;所有 MLflow 运行都属于试验。Experiments are the primary unit of organization in MLflow; all MLflow runs belong to an experiment. 每个试验都允许可视化、搜索和比较运行,以及下载运行项目或元数据以便在其他工具中进行分析。Each experiment lets you visualize, search, and compare runs, as well as download run artifacts or metadata for analysis in other tools. 试验在 Azure Databricks 托管的 MLflow 跟踪服务器中进行维护。Experiments are maintained in an Azure Databricks hosted MLflow tracking server.

试验位于工作区文件树中。Experiments are located in the Workspace file tree. 可使用与管理其他工作区对象(如文件夹、笔记本和库)相同的工具来管理试验。You manage experiments using the same tools you use to manage other workspace objects such as folders, notebooks, and libraries.

笔记本Notebooks

下面的快速入门笔记本演示了如何使用 MLflow 跟踪 API 创建并记录 MLflow 运行,以及如何使用试验 UI 来查看运行。The following quick start notebooks demonstrate how to create and log to an MLflow run using the MLflow tracking APIs, as well how to use the experiment UI to view the run. 这些笔记本可在 Python、Scala 和 R 中使用。These notebooks are available in Python, Scala, and R.

Python 和 R 笔记本使用笔记本试验The Python and R notebooks use a notebook experiment. Scala 笔记本在 Shared 文件夹中创建试验。The Scala notebook creates an experiment in the Shared folder.