def connect(self, trainer: 'pl.Trainer', *args: Any, **kwargs: Any) -> None:
"""Connects the loop with all necessary parts like trainer and accelerators"""
super().connect(trainer, *args, **kwargs)
self.batch_loop = TrainingBatchLoop()
self.batch_loop.connect(trainer)
self.val_loop = loops.EvaluationLoop()
self.val_loop.connect(trainer)
def __init__(self,
min_steps: Optional[int] = None,
max_steps: int = -1) -> None:
super().__init__()
if max_steps is None:
rank_zero_deprecation(
"Setting `max_steps = None` is deprecated in v1.5 and will no longer be supported in v1.7."
" Use `max_steps = -1` instead.")
max_steps = -1
elif max_steps < -1:
raise MisconfigurationException(
f"`max_steps` must be a non-negative integer or -1 (infinite steps). You passed in {max_steps}."
)
self.min_steps = min_steps
self.max_steps = max_steps
self.batch_progress = BatchProgress()
self.scheduler_progress = SchedulerProgress()
self.batch_loop = TrainingBatchLoop()
self.val_loop = loops.EvaluationLoop(verbose=False)
self._results = _ResultCollection(training=True)
self._outputs: _OUTPUTS_TYPE = []
self._warning_cache = WarningCache()
# caches the loaded dataloader state until dataloader objects are available
self._dataloader_state_dict: Dict[str, Any] = {}
self._batches_that_stepped: int = 0
def __init__(self, min_steps: int, max_steps: int):
super().__init__()
self.min_steps: int = min_steps
self.max_steps: int = max_steps
self.global_step: int = 0
# the total batch index across all epochs
self.total_batch_idx: int = 0
# the current batch index in the loop that runs over the dataloader(s)
self.iteration_count: int = 0
# the current split index when the batch gets split into chunks in truncated backprop through time
self.split_idx: Optional[int] = None
# the number of batches seen this run, updates immediately after batch_loop.run()
self.batches_seen: int = 0
self.is_last_batch: Optional[bool] = None
self.progress = TrainingEpochProgress()
self.batch_loop = TrainingBatchLoop()
self.val_loop = loops.EvaluationLoop()
self._results = ResultCollection(training=True)
self._dataloader_idx: Optional[int] = None
self._warning_cache: WarningCache = WarningCache()
self._epoch_output: Optional[List[List[STEP_OUTPUT]]] = None