Rename leftover definitions in Lite tests (#16309)

Author: awaelchli

src/pytorch_lightning/accelerators/tpu.py

@@ -16,7 +16,7 @@
 import torch
 
 from lightning_fabric.accelerators.tpu import _parse_tpu_devices, _XLA_AVAILABLE
-from lightning_fabric.accelerators.tpu import TPUAccelerator as LiteTPUAccelerator
+from lightning_fabric.accelerators.tpu import TPUAccelerator as FabricTPUAccelerator
 from lightning_fabric.utilities.types import _DEVICE
 from pytorch_lightning.accelerators.accelerator import Accelerator
 
@@ -74,7 +74,7 @@ def auto_device_count() -> int:
 
     @staticmethod
     def is_available() -> bool:
-        return LiteTPUAccelerator.is_available()
+        return FabricTPUAccelerator.is_available()
 
     @classmethod
     def register_accelerators(cls, accelerator_registry: Dict) -> None:

src/pytorch_lightning/plugins/precision/precision_plugin.py

@@ -21,13 +21,13 @@
 from torch.optim import Optimizer
 
 import pytorch_lightning as pl
-from lightning_fabric.plugins import Precision as LitePrecision
+from lightning_fabric.plugins import Precision as FabricPrecision
 from lightning_fabric.utilities.types import Steppable
 from pytorch_lightning.core.hooks import CheckpointHooks
 from pytorch_lightning.utilities import grad_norm, GradClipAlgorithmType
 
 
-class PrecisionPlugin(LitePrecision, CheckpointHooks):
+class PrecisionPlugin(FabricPrecision, CheckpointHooks):
     """Base class for all plugins handling the precision-specific parts of the training.
 
     The class attribute precision must be overwritten in child classes. The default value reflects fp32 training.

src/pytorch_lightning/strategies/colossalai.py

@@ -406,7 +406,7 @@ def optimizer_step(
         **kwargs: Any,
     ) -> Any:
         model = model or self.lightning_module
-        # TODO(lite): remove assertion once strategy's optimizer_step typing is fixed
+        # TODO(fabric): remove assertion once strategy's optimizer_step typing is fixed
         assert isinstance(model, pl.LightningModule)
         return self.precision_plugin.optimizer_step(
             optimizer, model=model, optimizer_idx=opt_idx, closure=closure, **kwargs

src/pytorch_lightning/strategies/deepspeed.py

@@ -608,7 +608,7 @@ def _format_config(self) -> None:
         self._format_precision_config()
 
     def _format_batch_size_and_grad_accum_config(self) -> None:
-        # todo: using lite, we do not support these variables within the config
+        # TODO: Using Fabric, we do not support these variables within the config
         assert isinstance(self.config, dict)
         if self.lightning_module is None:
             return

src/pytorch_lightning/strategies/strategy.py

@@ -229,7 +229,7 @@ def optimizer_step(
             \**kwargs: Keyword arguments to to ``optimizer.step``
         """
         model = model or self.lightning_module
-        # TODO(lite): remove assertion once strategy's optimizer_step typing is fixed
+        # TODO(fabric): remove assertion once strategy's optimizer_step typing is fixed
         assert isinstance(model, pl.LightningModule)
         return self.precision_plugin.optimizer_step(
             optimizer, model=model, optimizer_idx=opt_idx, closure=closure, **kwargs
@@ -241,19 +241,19 @@ def _setup_model_and_optimizers(self, model: Module, optimizers: List[Optimizer]
         The returned objects are expected to be in the same order they were passed in. The default implementation will
         call :meth:`_setup_model` and :meth:`_setup_optimizer` on the inputs.
         """
-        # TODO: standardize this across all plugins in Lightning and Lite. Related refactor: #7324
+        # TODO: standardize this across all plugins in Lightning and Fabric. Related refactor: #7324
         model = self._setup_model(model)
         optimizers = [self._setup_optimizer(optimizer) for optimizer in optimizers]
         return model, optimizers
 
     def _setup_model(self, model: Module) -> Module:
         """Performs setup for the model, e.g., by wrapping it by another class."""
-        # TODO: standardize this across all plugins in Lightning and Lite. Related refactor: #7324
+        # TODO: standardize this across all plugins in Lightning and Fabric. Related refactor: #7324
         return model
 
     def _setup_optimizer(self, optimizer: Optimizer) -> Optimizer:
         """Performs setup for the optimizer, e.g., by wrapping it by another class."""
-        # TODO: standardize this across all plugins in Lightning and Lite. Related refactor: #7324
+        # TODO: standardize this across all plugins in Lightning and Fabric. Related refactor: #7324
         return optimizer
 
     def batch_to_device(self, batch: Any, device: Optional[torch.device] = None, dataloader_idx: int = 0) -> Any:

src/pytorch_lightning/strategies/utils.py

@@ -34,7 +34,7 @@ def on_colab_kaggle() -> bool:
 
 
 def _call_register_strategies(registry: _StrategyRegistry, base_module: str) -> None:
-    # TODO(lite): Remove this function once PL strategies inherit from Lite's Strategy base class
+    # TODO(fabric): Remove this function once PL strategies inherit from Fabrics Strategy base class
     module = importlib.import_module(base_module)
     for _, mod in getmembers(module, isclass):
         if issubclass(mod, Strategy) and _is_register_method_overridden(mod, Strategy, "register_strategies"):

src/pytorch_lightning/trainer/connectors/accelerator_connector.py

@@ -684,7 +684,7 @@ def _init_strategy(self) -> None:
         if isinstance(self._strategy_flag, str):
             self.strategy = StrategyRegistry.get(self._strategy_flag)
         elif isinstance(self._strategy_flag, Strategy):
-            # TODO(lite): remove ignore after merging lite and PL strategies
+            # TODO(fabric): remove ignore after merging Fabric and PL strategies
             self.strategy = self._strategy_flag  # type: ignore[assignment]
         else:
             raise RuntimeError(f"{self.strategy} is not valid type: {self.strategy}")

src/pytorch_lightning/trainer/trainer.py

@@ -1681,7 +1681,7 @@ def accelerator(self) -> Accelerator:
 
     @property
     def strategy(self) -> Strategy:
-        # TODO(lite): remove ignore after merging lite and PL strategies
+        # TODO(fabric): remove ignore after merging Fabric and PL strategies
         return self._accelerator_connector.strategy  # type: ignore[return-value]
 
     @property

src/pytorch_lightning/utilities/imports.py

@@ -21,7 +21,7 @@
 _PYTHON_GREATER_EQUAL_3_8_0 = (sys.version_info.major, sys.version_info.minor) >= (3, 8)
 _PYTHON_GREATER_EQUAL_3_10_0 = (sys.version_info.major, sys.version_info.minor) >= (3, 10)
 _TORCH_LESSER_EQUAL_1_10_2 = compare_version("torch", operator.le, "1.10.2")
-# duplicated from lite because HPU is patching it below
+# duplicated from fabric because HPU is patching it below
 _TORCH_GREATER_EQUAL_1_13 = compare_version("torch", operator.ge, "1.13.0")
 
 _HABANA_FRAMEWORK_AVAILABLE = package_available("habana_frameworks")

tests/tests_fabric/accelerators/test_tpu.py

@@ -29,11 +29,11 @@ def test_availability():
 @pytest.mark.parametrize(
     "devices,expected",
     [
-        (0, []),  # TODO(lite): This should raise an exception
+        (0, []),  # TODO(fabric): This should raise an exception
         (1, [0]),
         (2, [0, 1]),
         (3, [0, 1, 2]),
-        ("anything-else", "anything-else"),  # TODO(lite): This should raise an exception
+        ("anything-else", "anything-else"),  # TODO(fabric): This should raise an exception
     ],
 )
 def test_get_parallel_devices(devices, expected):