feat: Add Neptune Scale logger support

README.md

@@ -279,6 +279,9 @@ trainer = Trainer(logger=loggers.MLFlowLogger())
 # neptune
 trainer = Trainer(logger=loggers.NeptuneLogger())
 
+# neptune scale
+trainer = Trainer(logger=loggers.NeptuneScaleLogger())
+
 # ... and dozens more
 ```
 

docs/source-pytorch/conf.py

@@ -367,6 +367,7 @@ def _load_py_module(name: str, location: str) -> ModuleType:
     # TODO: these are missing objects.inv
     # "comet_ml": ("https://www.comet.com/docs/v2/", None),
     # "neptune": ("https://docs.neptune.ai/", None),
+    # "neptune scale": ("https://docs-beta.neptune.ai/", None),
     # "wandb": ("https://docs.wandb.ai//", None),
 }
 nitpicky = True
@@ -476,6 +477,7 @@ def _load_py_module(name: str, location: str) -> ModuleType:
     ("py:meth", "move_data_to_device"),
     ("py:class", "neptune.Run"),
     ("py:class", "neptune.handler.Handler"),
+    ("py:class", "neptune_scale.Run"),
     ("py:meth", "on_after_batch_transfer"),
     ("py:meth", "on_before_batch_transfer"),
     ("py:meth", "on_save_checkpoint"),
@@ -620,7 +622,7 @@ def package_list_from_file(file):
 from lightning.pytorch.cli import _JSONARGPARSE_SIGNATURES_AVAILABLE as _JSONARGPARSE_AVAILABLE
 from lightning.pytorch.utilities.imports import _TORCHVISION_AVAILABLE
 from lightning.fabric.loggers.tensorboard import _TENSORBOARD_AVAILABLE, _TENSORBOARDX_AVAILABLE
-from lightning.pytorch.loggers.neptune import _NEPTUNE_AVAILABLE
+from lightning.pytorch.loggers.neptune import _NEPTUNE_AVAILABLE, _NEPTUNE_SCALE_AVAILABLE
 from lightning.pytorch.loggers.comet import _COMET_AVAILABLE
 from lightning.pytorch.loggers.mlflow import _MLFLOW_AVAILABLE
 from lightning.pytorch.loggers.wandb import _WANDB_AVAILABLE

docs/source-pytorch/visualize/supported_exp_managers.rst

@@ -101,6 +101,43 @@ Here's the full documentation for the :class:`~lightning.pytorch.loggers.Neptune
 
 ----
 
+Neptune Scale
+==========
+To use `Neptune Scale <https://docs-beta.neptune.ai/>`_ first install the neptune-scale package:
+
+.. code-block:: bash
+
+    pip install neptune-scale
+
+
+Configure the logger and pass it to the :class:`~lightning.pytorch.trainer.trainer.Trainer`:
+
+.. testcode::
+    :skipif: not _NEPTUNE_SCALE_AVAILABLE
+
+    from neptune_scale import Run
+    from lightning.pytorch.loggers import NeptuneScaleLogger
+
+    neptune_scale_logger = NeptuneScaleLogger(
+        api_key=<YOUR_NEPTUNE_SCALE_API_KEY>,  # replace with your own
+        project="common/pytorch-lightning-integration",  # format "<WORKSPACE/PROJECT>"
+    )
+    trainer = Trainer(logger=neptune_scale_logger)
+
+Access the Neptune Scale logger from any function (except the LightningModule *init*) to use its API for tracking advanced artifacts
+
+.. code-block:: python
+
+    class LitModel(LightningModule):
+        def any_lightning_module_function_or_hook(self):
+            neptune_scale_logger = self.logger.experiment
+            neptune_scale_logger.log_metrics(data={"path/to/metadata": metadata}, step=step)
+            neptune_scale_logger.log_configs(data={"path/to/config": config})
+
+Here's the full documentation for the :class:`~lightning.pytorch.loggers.NeptuneScaleLogger`.
+
+----
+
 Tensorboard
 ===========
 `TensorBoard <https://pytorch.org/docs/stable/tensorboard.html>`_ can be installed with:

requirements/pytorch/loggers.info

@@ -1,6 +1,7 @@
 # all supported loggers. this list is here as a reference, but they are not installed in CI
 
 neptune >=1.0.0
+neptune-scale
 comet-ml >=3.31.0
 mlflow >=1.0.0
 wandb >=0.12.10

src/lightning/pytorch/CHANGELOG.md

@@ -27,6 +27,15 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 -
 
 
+---
+
+## [unreleased] - 2025-03-31
+
+### Added
+
+- Add support for Neptune Scale logger ([#20686](https://github.com/Lightning-AI/pytorch-lightning/pull/20686))
+
+
 ---
 
 ## [2.5.1] - 2025-03-18

src/lightning/pytorch/loggers/__init__.py

@@ -15,8 +15,17 @@
 from lightning.pytorch.loggers.csv_logs import CSVLogger
 from lightning.pytorch.loggers.logger import Logger
 from lightning.pytorch.loggers.mlflow import MLFlowLogger
-from lightning.pytorch.loggers.neptune import NeptuneLogger
+from lightning.pytorch.loggers.neptune import NeptuneLogger, NeptuneScaleLogger
 from lightning.pytorch.loggers.tensorboard import TensorBoardLogger
 from lightning.pytorch.loggers.wandb import WandbLogger
 
-__all__ = ["CometLogger", "CSVLogger", "Logger", "MLFlowLogger", "TensorBoardLogger", "WandbLogger", "NeptuneLogger"]
+__all__ = [
+    "CometLogger",
+    "CSVLogger",
+    "Logger",
+    "MLFlowLogger",
+    "TensorBoardLogger",
+    "WandbLogger",
+    "NeptuneLogger",
+    "NeptuneScaleLogger",
+]