Dominodatalab

Added

Changed
* Updated authentication methods precedence.
* Changed version number message logging level to debug

Added
* Added support for MPI cluster type in job_start
* Added support for overriding the default hardware tier by id when starting a job

Changed
* Overriding the hardware tier by name has been marked as deprecated

Added

* Automatically re-read auth token from file to avoid reusing short-lived access tokens
* Better token expiration error handling for job_start and runs_start methods
* Added support for external volume mounts in job_start

Changed

Added

* Added support for launching jobs with Dask clusters via job_start
* Added the ability to choose which exceptions to ignore (if any) while polling for `job_start_blocking`
* Added several new unit tests in `test_basic_auth.py` and `test_jobs.py`
* Added a public method to re-authenticate if a token expires (assuming a long-running process)
* Allow `auth_token` to be passed as a string to the `__init__()` method

Changed

Added

* `collaborators_remove` method
* Model export and AWS SageMaker API integration

Changed

* Model version publish method (model_version_publish) doesn't take name anymore
* `collaborators_add` method changed to use v4 API

Added

* Add v4 API endpoint `job_start` to start job with `onDemandSparkCluster` option
* Add v4 API endpoint `job_stop`, `job_status` & `job_start_blocking`

Changed

* Started caching project_id
* Started using request session instead of creating new session every time
* Check if an App is running before attempting to stop it (in `app_unpublish`)