- Bugfix
transitions update-executionnow supports additional keyword arguments - Added optional parameter
--asynctopredictions create - Added
predictions get - Added optional parameter
--statistics-last-n-daystomodels get-training - Added optional parameter
--max-elementstomodels get-training - Remove optional parameter
--in-schemaand--out-schemafromtransitions createandtransitions update
- Bugfix
models update-trainingnow works as intended when specifying--deployment-environment-id
- Bugfix
models update-trainingno longer attempts to sendnullformetadatawhen not specified
- Added optional parameter
--statustoworkflows update
- Added optional parameter
--qualitytodocuments get
- Added
app-clients get - Added optional parameter
--statustoworkflows update-execution - Added optional parameter
--next-transition-idtoworkflows update-execution - Removed mandatory parameter
next_transition_idfromworkflows update-execution - Added optional parameter
--statistics-last-n-daystomodels get
- Updated default
--preprocess-imageincreate-defaultworkflow
- Added optional parameter
--email-configtoworkflows create - Added optional parameter
--email-configtoworkflows update - Now able to also specify inline JSON for optional parameter
--error-configinworkflows create - Now able to also specify inline JSON for optional parameter
--error-configinworkflows update - Now able to also specify inline JSON for optional parameter
--completed-configinworkflows create - Now able to also specify inline JSON for optional parameter
--completed-configinworkflows update
- Added
models get-data-bundle - Added
models get-training
- Updated
--create-default-workflowinworkflows create-defaultto use the latest docker images and automatically addmodelIdto the form config
- Added
datasets create-transformation - Added
datasets delete-transformation - Added
datasets list-transformations - Added Python 3.11 support
- Removed Python 3.6 and 3.7 support as they have reached end of life
- Added optional parameter
--model-idtopredictions list
- Added optional parameter
--metadatatoworkflows create - Added optional parameter
--metadatatoworkflows update
- Added
roles list - Added
roles get - Added optional parameter
--role-idstoapp-clients create - Added optional parameter
--role-idstoapp-clients update - Added optional parameter
--role-idstousers create - Added optional parameter
--role-idstousers update - Removed deprecated optional parameters
--avatarand--namefromusers create - Removed deprecated optional parameters
--avatarand--namefromusers update
- Added optional parameter
--widthtodocuments get - Added optional parameter
--heighttodocuments get - Added optional parameter
--densitytodocuments get - Added optional parameter
--pagetodocuments get - Added optional parameter
--rotationtodocuments get
- Added
deployment-environments get - Added
deployment-environments list - Added
--postprocess-configtomodels create - Updated
--preprocess-configinmodels createto also accept inline JSON - Added
--postprocess-configtomodels update - Updated
--preprocess-configinmodels updateto also accept inline JSON - Updated
--postprocess-configinpredictions createto also accept path to a JSON file
- Added
--rotationtopredictions create
- Added
--base-modeltomodels create - Added
--ownertomodels list
- Fixed a bug causing
datasets create-documentsto skip creating new documents for documents that have been deleted from the dataset since last upload.
- Updated
workflows create-defaultto use the default manual transition form component - Updated default
--preprocess-imageinworkflows create-default
- Fixed a bug in
datasets create-documents.
- Using a CSV file as input to
datasets create-documentswill now try to locate the document file specified bydocument-path-columnin all sub-directories, e.g. ifdocument_path=foo/bar/baz.pdfit will first see iffoo/bar/baz.pdfpoints to an existing and valid file type, then it will check ifbar/baz.pdfpoints to an existing and valid file type and so forth. - Fixed output bug in
datasets get-documents. Now only counts number of documents from the dataset you are downloading, not the number of documents in the outdir_dir
- Added
--image-urltotransitions update - Added
--secret-idtotransitions update - Added
--cputotransitions update - Added
--memorytotransitions update - Now supporting
transitions update --environment nullto clear all environment variables from a transition - Now supporting
transitions update --environment-secrets nullto clear all environment secrets from a transition - Added
--from-start-timetoworkflows list-executions - Added
--to-start-timetoworkflows list-executions - Added shorthand
-pfor--profile
- Added CSV support in
datasets create-documents - Removed deprecated
--widthand--heightfrommodels createandmodels update
- You may now upload documents without ground truth using
datasets create-documents - Improved the caching mechanism of
datasets create-documents
- Added
originalFilePathtometadataindatasets create-documents
- Fix bug in datasets create-documents
- Added optional parameter
--use-cachetodatasets create-documents - Don't use cached data by default in
datasets create-documents
- Reduced memory overhead in
datasets create-documents
- Optional parameter
--training-idinmodels updateis now nullable
- Added
payment-methods create - Added
payment-methods delete - Added
payment-methods get - Added
payment-methods list - Added
payment-methods update - Added optional parameter
--payment-method-idtoorganizations update
- Added optional parameter
--sort-bytodocuments list - Added optional parameter
--sort-bytopredictions list - Added optional parameter
--ordertodocuments list - Added optional parameter
--ordertopredictions list
- Added
datasets get-documents. This command will download all documents (PDF, JPEG etc. and JSON ground truth) to directory specified byoutput_dir - Show help when invoking
laswith no arguments - Support newer Python versions in setup.py
- Added --version
- Fix bug in datasets create-documents
- width and height is now optional in models create
- Added --training-id to predictions create
- Added --training-id to models update
- Support csv files in datasets create-documents
- Added models list-all-data-bundles
- Added optional parameter --metadata to documents create, documents update, datasets create, datasets update, models create, models update and trainings create
- Removed optional parameter --start-training from models update (Breaking change)
- Changed several optional argument names from --X-path to --X (Breaking change)
- Added models update-training
- Update argcomplete package
- Added exception handling to datasets create-documents
- Support specifying ground-truth encoding in datasets create-documents
- Added models list-trainings
- Added models create-training
- Added plans list
- Added plans get
- Added optional parameter --postprocess-config to predictions create
- Removed all support for batches. Use datasets instead.
- Renamed datasets sync to datasets create-documents and support folder as input and better help functionality
- Updated version of Python SDK to 4.0.0. API key is no longer needed.
- Bugfix in datasets sync
- Added datasets sync for uploading larger datasets
- Added datasets create
- Added datasets list
- Added datasets update
- Added datasets delete
- Added optional parameter --dataset-id to documents create
- Added optional parameter --document-id to documents delete
- Added optional parameter --dataset-id to documents create
- Added optional parameter --dataset-id to documents list
- Added optional parameter --dataset-id to documents delete
- Added optional parameter --dataset-id to documents update
- Added models create-data-bundle
- Added models list-data-bundles
- Added models update-data-bundle
- Added models delete-data-bundle
las documents delete will now require a document id as a positional argument and delete that document only.
las documents delete-all can be used to delete multiple documents. It will by default delete all the documents available.
Use --max-results to restrict the number of documents deleted as before,
but the default will be to delete everything.
- Fixed bug causing more log messages than necessary to be printed
- Added organizations get
- Added organizations update
- Added --login-urls and --default-login-url to app-clients create
- Added models create
- Added models update
- Added models get
- Added batches update
- Added app-clients update
- Added argument app_client_id to users create