1 9 Devel - Pytorch Lightning

.. list-table:: devel 1.9 :widths: 40 40 20 :header-rows: 1

- If
- Then
- Ref
- passed the pl_module argument to distributed module wrappers
- passed the (required) forward_module argument
- PR16386_
- used DataParallel and the LightningParallelModule wrapper
- use DDP or DeepSpeed instead
- PR16748_ :doc:DDP <../../accelerators/gpu_expert>
- used pl_module argument from the distributed module wrappers
- use DDP or DeepSpeed instead
- PR16386_ :doc:DDP <../../accelerators/gpu_expert>
- called pl.overrides.base.unwrap_lightning_module function
- use DDP or DeepSpeed instead
- PR16386_ :doc:DDP <../../accelerators/gpu_expert>
- used or derived from pl.overrides.distributed.LightningDistributedModule class
- use DDP instead
- PR16386_ :doc:DDP <../../accelerators/gpu_expert>
- used the pl.plugins.ApexMixedPrecisionPlugin plugin
- use PyTorch native mixed precision
- PR16039_
- used the pl.plugins.NativeMixedPrecisionPlugin plugin
- switch to the pl.plugins.MixedPrecisionPlugin plugin
- PR16039_
- used the fit_loop.min_steps setters
- implement your training loop with Fabric
- PR16803_
- used the fit_loop.max_steps setters
- implement your training loop with Fabric
- PR16803_
- used the data_parallel attribute in Trainer
- check the same using isinstance(trainer.strategy, ParallelStrategy)
- PR16703_
- used any function from pl.utilities.xla_device
- switch to pl.accelerators.XLAAccelerator.is_available()
- PR14514_ PR14550_
- imported functions from pl.utilities.device_parser.*
- import them from lightning_fabric.utilities.device_parser.*
- PR14492_ PR14753_
- imported functions from pl.utilities.cloud_io.*
- import them from lightning_fabric.utilities.cloud_io.*
- PR14515_
- imported functions from pl.utilities.apply_func.*
- import them from lightning_utilities.core.apply_func.*
- PR14516_ PR14537_
- used any code from pl.core.mixins
- use the base classes
- PR16424_
- used any code from pl.utilities.distributed
- rely on Pytorch's native functions
- PR16390_
- used any code from pl.utilities.data
- it was removed
- PR16440_
- used any code from pl.utilities.optimizer
- it was removed
- PR16439_
- used any code from pl.utilities.seed
- it was removed
- PR16422_
- were using truncated backpropagation through time (TBPTT) with LightningModule.truncated_bptt_steps
- use manual optimization
- PR16172_ :doc:Manual Optimization <../../model/manual_optimization>
- were using truncated backpropagation through time (TBPTT) with LightningModule.tbptt_split_batch
- use manual optimization
- PR16172_ :doc:Manual Optimization <../../model/manual_optimization>
- were using truncated backpropagation through time (TBPTT) and passing hidden to LightningModule.training_step
- use manual optimization
- PR16172_ :doc:Manual Optimization <../../model/manual_optimization>
- used pl.utilities.finite_checks.print_nan_gradients function
- it was removed
- used pl.utilities.finite_checks.detect_nan_parameters function
- it was removed
- used pl.utilities.parsing.flatten_dict function
- it was removed
- used pl.utilities.metrics.metrics_to_scalars function
- it was removed
- used pl.utilities.memory.get_model_size_mb function
- it was removed
- used pl.strategies.utils.on_colab_kaggle function
- it was removed
- PR16437_
- used LightningDataModule.add_argparse_args() method
- switch to using LightningCLI
- PR16708_
- used LightningDataModule.parse_argparser() method
- switch to using LightningCLI
- PR16708_
- used LightningDataModule.from_argparse_args() method
- switch to using LightningCLI
- PR16708_
- used LightningDataModule.get_init_arguments_and_types() method
- switch to using LightningCLI
- PR16708_
- used Trainer.default_attributes() method
- switch to using LightningCLI
- PR16708_
- used Trainer.from_argparse_args() method
- switch to using LightningCLI
- PR16708_
- used Trainer.parse_argparser() method
- switch to using LightningCLI
- PR16708_
- used Trainer.match_env_arguments() method
- switch to using LightningCLI
- PR16708_
- used Trainer.add_argparse_args() method
- switch to using LightningCLI
- PR16708_
- used pl.utilities.argparse.from_argparse_args() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.argparse.parse_argparser() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.argparseparse_env_variables() function
- switch to using LightningCLI
- PR16708_
- used get_init_arguments_and_types() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.argparse.add_argparse_args() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.parsing.str_to_bool() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.parsing.str_to_bool_or_int() function
- switch to using LightningCLI
- PR16708_
- used pl.utilities.parsing.str_to_bool_or_str() function
- switch to using LightningCLI
- PR16708_
- derived from pl.utilities.distributed.AllGatherGrad class
- switch to PyTorch native equivalent
- PR15364_
- used PL_RECONCILE_PROCESS=1 env. variable
- customize your logger
- PR16204_
- if you derived from mixin’s method pl.core.saving.ModelIO.load_from_checkpoint
- rely on pl.core.module.LightningModule
- PR16999_
- used Accelerator.setup_environment method
- switch to Accelerator.setup_device
- PR16436_
- used PL_FAULT_TOLERANT_TRAINING env. variable
- implement own logic with Fabric
- PR16516_ PR16533_
- used or derived from public pl.overrides.distributed.IndexBatchSamplerWrapper class
- it is set as protected
- PR16826_
- used the DataLoaderLoop class
- use manual optimization
- PR16726_ :doc:Manual Optimization <../../model/manual_optimization>
- used the EvaluationEpochLoop class
- use manual optimization
- PR16726_ :doc:Manual Optimization <../../model/manual_optimization>
- used the PredictionEpochLoop class
- use manual optimization
- PR16726_ :doc:Manual Optimization <../../model/manual_optimization>
- used trainer.reset_*_dataloader() methods
- use Loop.setup_data() for the top-level loops
- PR16726_
- used LightningModule.precision attribute
- rely on Trainer precision attribute
- PR16203_
- used Trainer.model setter
- you shall pass the model in fit/test/predict method
- PR16462_
- relied on pl.utilities.supporters.CombinedLoaderIterator class
- pass dataloders directly
- PR16714_
- relied on pl.utilities.supporters.CombinedLoaderIterator class
- pass dataloders directly
- PR16714_
- used pl.callbacks.progress.base.ProgressBarBase
- rename to pl.callbacks.progress.ProgressBar
- PR17058_
- accessed ProgressBarBase.train_batch_idx property
- rely on Trainer internal loops’ properties
- PR16760_
- accessed ProgressBarBase.val_batch_idx property
- rely on Trainer internal loops’ properties
- PR16760_
- accessed ProgressBarBase.test_batch_idx property
- rely on Trainer internal loops’ properties
- PR16760_
- accessed ProgressBarBase.predict_batch_idx property
- rely on Trainer internal loops’ properties
- PR16760_
- used Trainer.prediction_writer_callbacks property
- rely on precision plugin
- PR16759_
- used PrecisionPlugin.dispatch
- it was removed
- PR16618_
- used Strategy.dispatch
- it was removed
- PR16618_