Patch Notes

Lots of bugfixes, and some changes to deployment prefixes.

• Bayesian optimization improvements and bugfixes

• Data management related bugfixes and improvements.

Small bugfixes and some deployment scrutability improvements.

• The triumphant return of the execution outputs endpoint!
• Some bugfixes.

Totally not April Fools!

• Notifications!
• Nicer commit referencing!
• Execution time limits!
• Deployment polish!
• … and other bugfixes

Socially isolated bugfixes! #COVID19

• Environment and deployment management improvements
• Bugfixes here and there

• Bayesian hyperparameter optimization!
• Deployment memory requests!
• Agent updates!

• Input selection embetterments!
• Random Search tasks!
• Less annoying notifications!

Thanks to the holidays and everything, not much more than usability fixes and bugfixes this time! A little polish never hurts, though, right?

• Execution Alerts!
• Bugfixes!

• Refreshed organization and profile views
• Bugfixes

• Improved onboarding experience
• A new template project by Hugging Face
• Little bugfixes

• Additional enterprise features like project ownership control
• API token management improvements
• Bug-swatting, as usual

• Enterprise goodness edition!
• Team Access Grants!
• Azure AD Login!

Usability and performance improvements, new CLI version.

• Better UX and more control over package contents in Jupyhai
• A cornucopia of bug fixes and optimizations, as well as Hard Stop for executions.

• Logs shown in your local time zone.
• Allow copying executions into tasks.
• Continued pipeline improvements like deletion and better error messages on invalid pipelines.
• And other minor improvements.

• Pipelines are in open beta now!
• Data heritage view!
• API call view!
• Scatter charts!
• A bunch of bugfixes!
• #hashtags!
• Exclamation marks!

• New CLI version!
• Jupyter Notebook previews!
• Robustness! Stability! Checksums!

• Some little bugfixes and a some quality-of-life improvements, plus some new features bubbling under.

• Some little bugfixes, a sprinkling little quality-of-life improvements, a handful of new features.
  What's not to like?

Just a small bugfix release to make your Friday a little brighter.

• Quality-of-life improvements
• Metadata, parameters and input information available in /valohai/config
• Input data tracking
• A bunch of small bugfixes
• Happy Midsummer!

• Improved execution log filtering using | syntax
• Improved and more accurate Valohai API documentation
• Remove length limit from project repository reference to better allow multiple branches in a project

• Support for enormous amounts of execution output files
• Support for execution output files with very long names
• Read-only and write-only data stores

• Support for new CUDA images (e.g. Tensorflow 1.13.x)
• Various platform Quality of Life improvements and bugfixes
• New Valohai command-line client version

• Ad-hoc execution copying
• Error codes in API responses

• New tutorial experience
• Commit handling improvements
• Metadata and parameter formatting and sorting fixes

• Private Docker images
• Metadata comparison
• Bugfixes and performance improvements

• Improved Swift support
• Quality-of-life improvements
• Minor bugfixes

• Execution and task re-titling
• Nicer input editing
• Repository URL buttons
• Deployment traceability
• Execution copy bugfix

• Task copying
• Execution table improvements
• Metadata chart improvements
• Team/organization membership improvements
• OpenSSH format key support

• Organization invitation system
• User experience improvements

• Style refresh
• Metadata chart improvements and bugfixes
• A plethora of other small quality of life improvements

• Superpowered chart views!
• Wildcards and URI resolution for custom data stores!
• OpenStack Swift data storage support!
• Bugfixes and polish!
• Happy holidays!

• Allow enforcing two-factor authentication.
• New project dashboard view.
• Technical environmental variable view.
• Native Google Storage support.

• Two-factor authentication with TOTP
• Allow writing notes to tasks
• Improved autoscaling logic

• Random hyperparameter optimization
• More user-friendly execution/task creation

• Improved execution tagging system that now also covers tasks.
• We provide suggestions for execution Docker images.
• Metadata charts now properly show zero values.

• Error codes from executions are now handled correctly.
• Small UI polish and quality-of-life improvements in many places.

• Built-in environment variable support.
• Step inputs can now have descriptions.
• Improved Create Execution and Create Task layouts.

• Organizations!
• Nicer invoices!
• Project transfers!
• Exclamation marks!

• Improvements to the execution view page.
• Executions and tasks can be titled.
• Repositories no longer require valohai.yaml.
• Billing is available in EUR too.
• New parameter and parameter-value interpolations in addition to parameters.
• More useful data is printed at start of execution.
• Valohai-CLI was updated.
• It's hot in Turku, Finland.

• Layout changes, project header should now be more clear on all projects.
• You can now easily preview image output files on web interface.
• Kubernetes deployment test view supports image responses for more streamlined testing.

• New usage and bill estimation page.
• Momentarily losing server connection won't hide logs on execution page.

• Execution metadata charts are lazy-loaded.
• Data browser will now correctly refresh available files after upload.
• Convert all frontend code to TypeScript.

• Add deployment endpoint test UI with optional drag and drop image payload.
• Internal crash reporting now includes user contact information.

• Non-integer number values for parameters and metadata are rounded.
• Allow ordering executions by parameter.
• Removed Promise.finally() usage that caused Linux browsers to crash on execution creation.

• Shell variable syntax ${foo} is now supported within commands.
• Stores can now be deleted.

• Improvements to metadata and log views.
• Custom S3 stores no longer require the Multipart IAM Upload Role field.
• Retry functionality in the frontend now works better.

• Kubernetes deployment feature in beta for Early Adopter users.
• Improve task detail view layout.
• Hide purged files when selecting execution inputs.

• Allow downloading execution logs through the web client.
• Show execution durations in more human-readable format.
• Command-line client properly hides deleted executions.

• Show parameters of individual executions in the execution list on task page of the web client.
• Rename execution "Discussion" feature to "Notes" to clarify what it should be used for.

• Allow specifying multiple input files per execution input.
• Allow searching input files by the execution number that generated the data.
• Improve the layout of execution and task lists on the web client.

• Use S3 multipart upload to allow output files up to 5 terabytes.
• Add logarithmic hyperparameter optimization.

• Don't upload output directory files that start with a dot.
• Allow executions to upload files in the middle of their run aka. "live upload."

• Make it more comfortable to select machine type on the command-line client.
• Add credit system to the billing.
• Improve accuracy of execution duration calculations.
• Empty lines on hyperparameter optimization parameters are now ignored.

• Add execution tagging system.
• Allow users to download files through web client data browser.
• Show which machine type was used for an execution in the web client.
• Uploading won't overwrite an already existing S3 files.

• Add GitHub App integration to streamline onboarding of new users.
• Allow users to delete executions and uploaded files through the web client.
• Using custom Docker image will now start on the first try.

• Disallow copying of ad-hoc executions.
• Improve error reporting on git repository fetch errors for support.
• Load long executions logs without lagging the web client.

• Release the example project for R.
• Limit how many previous commit versions to show in execution create page dropdown.
• Avoid executions from entering eternal stopping limbo.

• Add new "Stopping" status for executions that take a long time to shut down.
• Allow project admins to set a default store, so it doesn't have to be defined before every execution and task.

• Allow hyperparameter optimization of flag parameter type.
• Custom stores can now be uploaded to.
• Limit how many log events are shown in the web client.
• Resign uploaded input file credentials before queueing an execution that uses then.

• Valohai platform is now out of beta!
• Add store system to allow customized download endpoints.
• Rework the usage flow how to define existing files as inputs to executions.

• Add user profile page.
• Rework how real-time statistics are displayed.

• Add support for non-GPU instances.
• Improve how fast execution status updates.
• Allow collaborators to see input and output files of projects they have been invited.

• Allow users to browse, download and manage their S3 output files on the web client.
• Allow users to use output and manually uploaded files as input without opening an S3 object to the public.

• Allow uploading input files to S3 stores with the web client.
• Allow short hash for commit selection in API and the command-line client.
• Implement per-user quota to restrict how much a single user can run in the shared clusters.

• Add full repository control to the REST API.
• Add more details about executions to the REST API.
• Add support for subdirectories in output upload.
• Input parameter 0 (zero) will now work correctly.

• Add support for writing notes on executions.
• Show links to documentation all around the web client so you can learn more about a specific feature of the system.
• Record real-time statistics like CPU or GPU so you can view the final values after completion.
• Non-US locale decimal point will now work with hyperparameter range definitions.

• Show total worker machine memory on real-time statistics.
• Now clicking usage number will take you to detail spending page.
• Allow importing tutorial projects even if you have completed the onboarding.

• Add pagination support for task executions list page.
• Show more detailed resource usage by project and by hardware type used for billing.
• Make worker infrastructure clean up cached content more rigorously.
• Navigate straight to the execution logs screen after execution creation.

• Improve project to repository linking user flow.
• Improve AWS cluster downscaling as workloads can drop from hundreds of parallel executions to just zero in mere minutes.
• Show more obvious error when the command-line client is used to start executions on a project without a linked repository.

• Release Valohai documentation website.
• Show billing estimate for the cloud resource usage.

• Add hyperparameter optimization and tasks.
• Allow project collaborators also see and use on-premises hardware.
• Add local directory mount as an input method for on-premises workers.

• Web client graphics are now served from a CDN for better speed.
• Stop canceled executions more gracefully.

• Allow users to select which type of AWS instances to run their execution on, not just g2.
• Stopped executions should not be picked again by Valohai workers.

• Make our AWS server clusters scale according to workloads we are having.
• Improve layout of the execution real-time statistics.
• Rework how browsers render execution logs; now with more color!

• Add support for project collaborators.
• New parameter type: flag.
• Now visitor will stay on the same execution screen tab when refreshing.
• Show real-time statistics of executions such as CPU and memory.

• Move default hosting of output files from the master server to S3.
• Allow users to cancel already started executions.
• Allow users to copy previous executions.
• Add support for optional execution inputs and parameters.
• Add support for multiple commands in a single execution.

• Parameters passed to the worker nodes are now properly quoted.

• Integrate Valohai YAML standard.
• Support for Valohai YAML steps, Docker image, command, inputs, and parameters.
• Add metadata support by reading JSON from standard output.
• Release the example project with TensorFlow and MNIST dataset.

• Release beta version of the Valohai machine learning platform for early access.
• Add login, billing, project, repository and execution management systems.