litlogger

Classes

LitLogger

Logger that enables remote experiment tracking, logging, and artifact management on lightning.ai.

Fabric/PyTorch Lightning logger that enables remote experiment tracking, logging, and artifact management on lightning.ai.

class lightning.pytorch.loggers.litlogger.LitLogger(root_dir=None, name=None, teamspace=None, metadata=None, store_step=True, log_model=False, save_logs=True, checkpoint_name=None)

Bases: Logger

Logger that enables remote experiment tracking, logging, and artifact management on lightning.ai.

Initialize the LightningLogger.

Parameters:

• root_dir (Union[str, Path, None]) – Folder where logs and metadata are stored (default: ./lightning_logs).

• name (Optional[str]) – Name of your experiment (defaults to a generated name).

• teamspace (Optional[str]) – Teamspace name where charts and artifacts will appear.

• metadata (Optional[dict[str, str]]) – Extra metadata to associate with the experiment as tags.

• log_model (bool) – If True, automatically log model checkpoints as artifacts.

• save_logs (bool) – If True, capture and upload terminal logs.

• checkpoint_name (Optional[str]) – Override the base name for logged checkpoints.

Example:

from lightning.pytorch import Trainer
from lightning.pytorch.demos.boring_classes import BoringModel, BoringDataModule
from lightning.pytorch.loggers.litlogger import LitLogger

class LoggingModel(BoringModel):
    def training_step(self, batch, batch_idx: int):
        loss = self.step(batch)
        # logging the computed loss
        self.log("train_loss", loss)
        return {"loss": loss}

trainer = Trainer(
    max_epochs=10,
    enable_model_summary=False,
    logger=LitLogger("./lightning_logs", name="boring_model")
)
model = BoringModel()
data_module = BoringDataModule()
trainer.fit(model, data_module)
trainer.test(model, data_module)

after_save_checkpoint(checkpoint_callback)

Called after a checkpoint is saved.

Logs checkpoints as artifacts if enabled.

Return type:

None

finalize(status=None)

Do any processing that is necessary to finalize an experiment.

Parameters:

status (Optional[str]) – Status that the experiment finished with (e.g. success, failed, aborted)

Return type:

None

get_file(path, verbose=True)

Download a file artifact from the cloud for this experiment.

Parameters:

• path (str) – Path where the file should be saved locally.

• verbose (bool) – Whether to print a confirmation message after download. Defaults to True.

Returns:

The local path where the file was saved.

Return type:

str

get_model(staging_dir=None, verbose=False, version=None)

Download and load a model object using litmodels.

Parameters:

• staging_dir (Optional[str]) – Optional directory where the model will be downloaded.

• verbose (bool) – Whether to show progress bar.

• version (Optional[str]) – Optional version string for the model.

Return type:

Any

Returns:

The loaded model object.

get_model_artifact(path, verbose=False, version=None)

Download a model artifact file or directory from cloud storage using litmodels.

Parameters:

• path (str) – Path where the model should be saved locally.

• verbose (bool) – Whether to show progress bar during download.

• version (Optional[str]) – Optional version string for the model.

Returns:

The local path where the model was saved.

Return type:

str

log_file(path)

Log a file as an artifact to the Lightning platform.

The file will be logged in the Teamspace drive, under a folder identified by the experiment name.

Parameters:

path (str) – Path to the file to log.

Return type:

None

Example::

logger = LitLogger(…) logger.log_file(‘config.yaml’)

log_graph(model, input_array=None)

Record model graph.

Parameters:

• model (Module) – the model with an implementation of forward.

• input_array (Optional[Tensor]) – input passes to model.forward

Return type:

None

log_hyperparams(params, metrics=None)

Log hyperparams.

Return type:

None

log_metadata(params)

Log hyperparams.

Return type:

None

log_metrics(metrics, step=None)

Records metrics. This method logs metrics as soon as it received them.

Parameters:

• metrics (Mapping[str, float]) – Dictionary with metric names as keys and measured quantities as values

• step (Optional[int]) – Step number at which the metrics should be recorded

Return type:

None

log_model(model, staging_dir=None, verbose=False, version=None, metadata=None)

Save and upload a model object to cloud storage.

Parameters:

• model (Any) – The model object to save and upload (e.g., torch.nn.Module).

• staging_dir (Optional[str]) – Optional local directory for staging the model before upload.

• verbose (bool) – Whether to show progress bar during upload.

• version (Optional[str]) – Optional version string for the model.

• metadata (Optional[dict[str, Any]]) – Optional metadata dictionary to store with the model.

Return type:

None

log_model_artifact(path, verbose=False, version=None)

Upload a model file or directory to cloud storage using litmodels.

Parameters:

• path (str) – Path to the local model file or directory to upload.

• verbose (bool) – Whether to show progress bar during upload. Defaults to False.

• version (Optional[str]) – Optional version string for the model. Defaults to the experiment version.

Return type:

None

save()

Save log data.

Return type:

None

property experiment: Optional[litlogger.Experiment]

Returns the underlying litlogger Experiment object.

property log_dir: str

The directory for this run’s tensorboard checkpoint.

By default, it is named 'version_${self.version}' but it can be overridden by passing a string value for the constructor’s version parameter instead of None or an int.

property name: str

Gets the name of the experiment.

property root_dir: str

Gets the save directory where the litlogger experiments are saved.

property sub_dir: Optional[str]

Gets the sub directory where the TensorBoard experiments are saved.

property version: Optional[str]

Get the experiment version - its time of creation.