deepaas-cli

Synopsis

deepaas-cli [options]

Description

deepaas-cli Command line interface (CLI) to DEEPaaS models

that are loaded through the deepaas.v2.models entrypoint API. One gets access to the same get_metadata, warm, predict, and train methods with all corresponding options as with DEEPaaS REST API. To get available for the method options, one has to call deepaas-cli <method> --help. Additional parameters are provided: --deepaas_method_output is to store the output in a pre-defined by a user file; --deepaas_with_multiprocessing is to activate multiprocessing support, default is True. If several models are available for loading, one has to provide which one to load via DEEPAAS_V2_MODEL environment setting.

Options

get_metadata

Calls get_metadata() method. The output can be stored via –deepaas_method_output.

warm

Calls warm() method, e.g. to prepare the model for execution.

predict

Calls predict() method. The output can be stored via –deepaas_method_output.

train

Calls train() method. The output can be stored via –deepaas_method_output.

--deepaas_method_output

To save the results to a local file, if needed. Available for get_metadata, predict, train methods.

--deepaas_with_multiprocessing

To activate multiprocessing support, default is True.

--model-name MODEL_NAME

Specify the model to be used. If not specified, DEEPaaS will fail if there are more than only one models available.

Files

None

See Also

Documentation: DEEPaaS API

Reporting Bugs

Bugs are managed at GitHub