Skip to main content

· 2 min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.8.0! MLflow 0.8.0 introduces several major features:

  • Dramatically improved UI for comparing experiment run results:

    • Metrics and parameters are by default grouped into a single column, to avoid an explosion of mostly-empty columns. Individual metrics and parameters can be moved into their own column to help compare across rows.
    • Runs that are "nested" inside other runs (e.g., as part of a hyperparameter search or multistep workflow) now show up grouped by their parent run, and can be expanded or collapsed altogether. Runs can be nested by calling mlflow.start_run or mlflow.run while already within a run.
    • Run names (as opposed to automatically generated run UUIDs) now show up instead of the run ID, making comparing runs in graphs easier.
    • The state of the run results table, including filters, sorting, and expanded rows, is persisted in browser local storage, making it easier to go back and forth between an individual run view and the table.

  • Support for deploying models as Docker containers directly to Azure Machine Learning Service Workspace (as opposed to the previously-recommended solution of Azure ML Workbench).

The release also includes bugfixes and improvements across the Python and Java clients, tracking UI, and documentation. Visit the change log to read more about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.7.0!

MLflow 0.7.0 introduces several major features:

The release also includes bugfixes and improvements across the Python and Java clients, tracking UI, and documentation. Visit the change log to read more about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.6.0!

MLflow 0.6.0 introduces several major features:

  • A Java client API (to be published on Maven within the next day or two)
  • Support for saving and serving SparkML models as MLeap for low-latency serving
  • Support for tagging runs with metadata, during and after the run completion
  • Support for deleting (and restoring deleted) experiments

In addition to these features, there are a host of improvements and bugfixes to the REST API, Python API, tracking UI, and documentation. Visit the change log to read more about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.5.2!

MLflow 0.5.2 is a patch release on top of 0.5.1 containing only bug fixes and no breaking changes or features.

Visit the change log to read about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.5.1!

MLflow 0.5.1 is a patch release on top of 0.5.0 containing only bug fixes and no breaking changes or features.

Visit the change log to read about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.5.0!

MLflow 0.5.0 offers some major improvements:

  • Keras and PyTorch first-class support as models
  • SFTP support as an artifactory
  • A new scatterplot visualization to compare runs
  • A more complete Python SDK for experiment and run management

Visit the change log to read about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.4.2!

MLflow 0.4.2 offers some improvements and minor bug fixes:

  • MLflow experiments REST API and mlflow experiments create now support providing --artifact-location
  • [UI] Runs can now be sorted by columns, and added a Select All button
  • Databricks File System (DBFS) artifactory support added
  • databricks-cli version upgraded to >= 0.8.0 to support new DatabricksConfigProvider interface

Visit the change log to read about the new features.

· One min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.4.1!

MLflow 0.4.1 offers some improvements and minor bug fixes:

  • [Projects] MLflow will use the conda installation directory given by the $MLFLOW_CONDA_HOME if specified (e.g. running conda commands by invoking "$MLFLOW_CONDA_HOME/bin/conda"), defaulting to running "conda" otherwise.
  • [UI] Show GitHub links in the UI for projects run from http(s):// GitHub URLs (#235, @smurching)

Visit the change log to read about the new features.

· 2 min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.4.0!

MLflow Release 0.4.0 is ready, released 2018-08-01. The release is available on PyPI and docs are updated. Here are the release notes (also available on GitHub):

Breaking changes:

  • [Projects] Removed the use_temp_cwd argument to mlflow.projects.run() (--new-dir flag in the mlflow run CLI). Runs of local projects now use the local project directory as their working directory. Git projects are still fetched into temporary directories (#215, @smurching)
  • [Tracking] GCS artifact storage is now a pluggable dependency (no longer installed by default). To enable GCS support, install google-cloud-storage on both the client and tracking server via pip (#202, @smurching).
  • [Tracking] Clients running MLflow 0.4.0 and above require a server running MLflow 0.4.0 or above, due to a fix that ensures clients no longer double-serialize JSON into strings when sending data to the server (#200, @aarondav). However, the MLflow 0.4.0 server remains backwards-compatible with older clients (#216, @aarondav)

Features:

  • [Examples] Add a more advanced tracking example: using MLflow with PyTorch and TensorBoard (#203)
  • [Models] H2O model support (#170, @ToonKBC)
  • [Projects] Support for running projects in subdirectories of Git repos (#153, @juntai-zheng)
  • [SageMaker] Support for specifying a compute specification when deploying to SageMaker (#185, @dbczumar)
  • [Server] Added --static-prefix option to serve UI from a specified prefix to MLflow UI and server (#116, @andrewmchen)
  • [Tracking] Azure blob storage support for artifacts (#206, @mateiz)
  • [Tracking] Add support for Databricks-backed RestStore (#200, @aarondav)
  • [UI] Enable productionizing frontend by adding CSRF support (#199, @aarondav)
  • [UI] Update metric and parameter filters to let users control column order (#186, @mateiz)

Bug fixes:

Visit the change log to read about the new features.

· 2 min read
MLflow maintainers

We are happy to announce the availability of MLflow 0.3.0!

MLflow Release 0.3.0 is ready, released 2018-07-18. The release is available on PyPI and docs are updated. Here are the release notes:

Breaking changes:

  • [MLflow Server] Renamed --artifact-root parameter to --default-artifact-root in mlflow server to better reflect its purpose (#165, @aarondav)

Features:

  • Spark MLlib integration: we now support logging SparkML Models directly in the log_model API, model format, and serving APIs (#72, @tomasatdatabricks)
  • Google Cloud Storage is now supported as an artifact storage root (#152, @bnekolny)
  • Support asychronous/parallel execution of MLflow runs (#82, @smurching)
  • [SageMaker] Support for deleting, updating applications deployed via SageMaker (#145, @dbczumar)
  • [SageMaker] Pushing the MLflow SageMaker container now includes the MLflow version that it was published with (#124, @sueann)
  • [SageMaker] Simplify parameters to SageMaker deploy by providing sane defaults (#126, @sueann)
  • [UI] One-element metrics are now displayed as a bar char (#118, @cryptexis)

Bug fixes:

Visit the change log to read about the new features.