MLflowVisBackend¶
- class mmengine.visualization.MLflowVisBackend(save_dir, exp_name=None, run_name=None, tags=None, params=None, tracking_uri=None, artifact_suffix=('.json', '.log', '.py', 'yaml'), tracked_config_keys=None)[source]¶
MLflow visualization backend class.
It can write images, config, scalars, etc. to a mlflow file.
Examples
>>> from mmengine.visualization import MLflowVisBackend >>> from mmengine import Config >>> import numpy as np >>> vis_backend = MLflowVisBackend(save_dir='temp_dir') >>> img = np.random.randint(0, 256, size=(10, 10, 3)) >>> vis_backend.add_image('img.png', img) >>> vis_backend.add_scalar('mAP', 0.6) >>> vis_backend.add_scalars({'loss': 0.1,'acc':0.8}) >>> cfg = Config(dict(a=1, b=dict(b1=[0, 1]))) >>> vis_backend.add_config(cfg)
- Parameters
save_dir (str) – The root directory to save the files produced by the backend.
exp_name (str, optional) – The experiment name. Defaults to None.
run_name (str, optional) – The run name. Defaults to None.
tags (dict, optional) – The tags to be added to the experiment. Defaults to None.
params (dict, optional) – The params to be added to the experiment. Defaults to None.
tracking_uri (str, optional) – The tracking uri. Defaults to None.
artifact_suffix (Tuple[str] or str, optional) – The artifact suffix. Defaults to (‘.json’, ‘.log’, ‘.py’, ‘yaml’).
tracked_config_keys (dict, optional) – The top level keys of config that will be added to the experiment. If it is None, which means all the config will be added. Defaults to None. New in version 0.7.4.
- add_config(config, **kwargs)[source]¶
Record the config to mlflow.
- Parameters
config (Config) – The Config object
- Return type
None
- add_scalar(name, value, step=0, **kwargs)[source]¶
Record the scalar data to mlflow.
- Parameters
name (str) – The scalar identifier.
value (int, float, torch.Tensor, np.ndarray) – Value to save.
step (int) – Global step value to record. Default to 0.
- Return type
None
- add_scalars(scalar_dict, step=0, file_path=None, **kwargs)[source]¶
Record the scalar’s data to mlflow.
- property experiment¶
Return MLflow object.