Configuration Options

The following is an overview of all available configuration options in Nova. For a sample configuration file, refer to Sample Configuration File.

DEFAULT

listen_ip
Type:

string

Default:

127.0.0.1

IP address on which the DEEPaaS API will listen.

The DEEPaaS API service listens on this IP address for incoming requests.

listen_port
Type:

port number

Default:

5000

Minimum Value:

0

Maximum Value:

65535

Port on which the DEEPaaS API will listen.

The DEEPaaS API service listens on this port number for incoming requests.

base_path
Type:

string

Default:

''

Base path for the API. This is useful when the API is served behind a reverse proxy that is not at the root of the domain. For example, if the API is served at https://example.com/deepaas, then the base path should be set to /deepaas. Defaults to the root of the domain.

train_endpoint
Type:

boolean

Default:

True

Specify whether DEEPaaS should provide a train endpoint (default: True).

predict_endpoint
Type:

boolean

Default:

True

Specify whether DEEPaaS should provide a predict endpoint (default: True).

debug_endpoint
Type:

boolean

Default:

False

Enable debug endpoint. If set we will provide all the information that you print to the standard output and error (i.e. stdout and stderr) through the “/debug” endpoint. Default is to not provide this information. This will not provide logging information about the API itself.

doc_endpoint
Type:

boolean

Default:

True

Enable documentation endpoint. If set we will provide the documentation through the “/api” endpoint. Default is to provide this information.

workers
Type:

integer

Default:

1

Specify the number of workers to spawn. If using a CPU you probably want to increase this number, if using a GPU probably you want to leave it to 1. (defaults to 1)

client_max_size
Type:

integer

Default:

0

Minimum Value:

0

Client’s maximum size in a request, in bytes. If a POST request exceeds this value, it raises an HTTPRequestEntityTooLarge exception. If set to 0, no file size limit will be enforced.

warm
Type:

boolean

Default:

True

Pre-warm the modules (eg. load models, do preliminary checks, etc). You might want to disable this option if DEEPaaS is loading more than one module because you risk getting out of memory errors.

model_name
Type:

string

Default:

''

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

debug
Type:

boolean

Default:

False

Enable debug mode in logging. This will provide more information about what is happening in the API. Default is to not provide this information.

log_level
Type:

string

Default:

INFO

Valid Values:

DEBUG, INFO, WARNING, ERROR, CRITICAL

Specify the log level for the API. Default is INFO. Possible values are DEBUG, INFO, WARNING, ERROR, CRITICAL.

log_file
Type:

string

Default:

''

Specify the log file to use. If not specified, logs will be sent to stdout.