finetuner.experiment module#

class finetuner.experiment.Experiment(client, name, status, created_at, description='')[source]#

Bases: object

Class for an experiment.

Parameters

  • client (FinetunerV1Client) – Client object for sending api requests.

  • name (str) – Name of the experiment.

  • status (str) – Status of the experiment.

  • created_at (str) – Creation time of the experiment.

  • description (Optional[str]) – Optional description of the experiment.

property name: str#
Return type

str

property status: str#
Return type

str

get_run(name)[source]#

Get a run by its name.

Parameters

name (str) – Name of the run.

Return type

Run

Returns

A Run object.

list_runs()[source]#

List every run inside the experiment.

Return type

List[Run]

Returns

List of Run objects.

delete_run(name)[source]#

Delete a run by its name.

Parameters

name (str) – Name of the run.

delete_runs()[source]#

Delete every run inside the experiment.

create_run(model, train_data, run_name=None, **kwargs)[source]#

Create a run inside the experiment.

Parameters

  • model (str) – Name of the model to be fine-tuned.

  • train_data (Union[DocumentArray, str]) – Either a DocumentArray for training data or a name of the DocumentArray that is pushed on Hubble.

  • run_name (Optional[str]) – Optional name of the run.

  • kwargs – Optional keyword arguments for the run config.

Return type

Run

Returns

A Run object.