mltk.core.SshMixin

class SshMixin

Provides various properties to the base MltkModel used by the ssh MLTK command.

See also

• Model Training via SSH

• Cloud Training with vast.ai

Properties

ssh_create_venv	If true then create the MLTK python virtual environment in the specified remote_directory.
ssh_download_files	List of remote files to download after invoking MLTK command.
ssh_environment	List of environment variables to export in the remote machine's shell session before invoking MLTK command.
ssh_remote_dir	Directory path to remote machine's workspace.
ssh_shutdown_cmds	List of shell commands to execute after invoking MLTK command.
ssh_startup_cmds	List of shell commands to execute before invoking MLTK command.
ssh_upload_files	List of local files to upload before invoking MLTK command.

Methods

__init__

property ssh_remote_dir

Directory path to remote machine’s workspace. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_startup_cmds

List of shell commands to execute before invoking MLTK command. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_shutdown_cmds

List of shell commands to execute after invoking MLTK command. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_upload_files

List of local files to upload before invoking MLTK command. All paths must be relative to the model specification script. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_download_files

List of remote files to download after invoking MLTK command. All paths must be relative to the model specification script. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_environment

List of environment variables to export in the remote machine’s shell session before invoking MLTK command. This overrides any settings in ~/.mltk/user_settings.yaml

See also

Model Training via SSH

property ssh_create_venv

If true then create the MLTK python virtual environment in the specified remote_directory. This overrides the setting in ~/.mltk/user_settings.yaml

See also

Model Training via SSH