experiment package¶

class experiment.experiment.Experiment(**kwargs)[source]¶

Bases: traitlets.config.application.Application

An experiment with CLI configuration support.

cache¶: An instance of a Python dict.

class_config_section(cls, classes=None)[source]¶

Get the config section for this class. :param classes: The list of other classes in the config file.

Used to reduce redundant information.

Parameters:	Note – I overwrite this function in order to have the configuration in “config.py” show the actual configuration used and have it non commented.

config_file¶: A trait for unicode strings.

config_file_name¶: A trait for unicode strings.

create_aliases()[source]¶: Flatten all class traits using aliasses.

custom_log_handlers¶: An instance of a Python list.

description¶: A trait for unicode strings.

exp_time¶

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

generate_config¶: A boolean (True, False) trait.

generate_config_file(classes=None)[source]¶: generate default config file from Configurables

initialize(argv=None)[source]¶

Do the basic steps to configure me.

Override in subclasses.

log_level¶: An enum whose value must be in a given sequence.

name¶: A trait for unicode strings.

results_path¶: A trait for unicode strings.

results_path_format¶: A trait for unicode strings.

run()[source]¶

The logic of the experiment.

Should be implemented by the subclass.

start()[source]¶: Start the whole thing

strict_git¶: A boolean (True, False) trait.

write_config()[source]¶: Write our config to a .py config file

class experiment.experiment.MLflowExperiment(**kwargs)[source]¶

Bases: experiment.experiment.Experiment

A singleton experiment with support for mlflow logging.

Note

This object will setup a connection with the mlflow server. To successfully do so it is recommended to setup the following environment variable:

MLFLOW_SERVER - URL:PORT of mlflow server.

description¶: A trait for unicode strings.

mlflow_server¶: A trait for unicode strings.

start()[source]¶: Start the whole thing

class experiment.experiment.VisdomExperiment(**kwargs)[source]¶

Bases: experiment.experiment.Experiment

A singleton experiment with support of Visdom logging.

Note

This object will setup a connection with the visdom server. To successfully do so it is recommended to setup the following environment variables:

VISDOM_SERVER_URL - URL of visdom server.
VISDOM_USERNAME - User name to use for logging to the visdom server (optional).
VISDOM_PASSWORD - Password to use for loffing to the visdom server (optional).

description¶: A trait for unicode strings.

vis¶

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

visdom_env¶: A trait for unicode strings.

visdom_params_win¶

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

visdom_server¶: A trait for unicode strings.

class experiment.experiment.TensorboardXExperiment(**kwargs)[source]¶

Bases: experiment.experiment.Experiment

A singleton experiment with support of TensorBoard logging.

Note

This object requires tensorBoardX to produce log summaries. The log summaries are stored in the path set by self.tb_log_dir. By default a unique folder will be created by script_name/jobid_date/ To successfully do so it is recommended to setup the following environment variable:

TENSORBOARD_BASE_DIR - Base path for storing log summaries (optional). Defaults to /tmp/tensorboard.

description¶: A trait for unicode strings.

log_to_text¶: A boolean (True, False) trait.

summary_writer¶

A trait whose value must be an instance of a specified class.

The value can also be an instance of a subclass of the specified class.

Subclasses can declare default classes by overriding the klass attribute

tb_log_dir¶: A trait for unicode strings.