* Added reference documentation. * Added a tutorial. * Added tests for the ML2P command line utilities. * Added support for attaching training and deployment instances to VPCs. * Open sourced ML2P under the ISCL.
0.1.5
------------------
* Added `ml2p dataset delete` which deletes an entire dataset. * Added `ml2p dataset ls` which lists the contents of a dataset. * Added `ml2p dataset up` which uploads a local file to a dataset. * Added `ml2p dataset dn` which downloads a file from a dataset. * Added `ml2p dataset rm` which deletes a file from a dataset.
0.1.4
------------------
* Correctly handle folder keys when downloading datasets from S3. Previously folder keys created files, now they created folders.
0.1.3
------------------
* Added support for local environments. These allow ML2P models to be trained and used to make predictions locally, as though they were being loaded in SageMaker. * Added support for downloading datasets and models from S3 into local environments.
0.1.2
------------------
* Fix support for recording predictions in S3 (in first release of this feature, the code attempted to pass a boolean value as an environment variable, which failed as expected).