update docs for "overfit_batches" (Lightning-AI#2324)

* update docs

* Apply suggestions from code review

Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>

Author: awaelchli

docs/source/debugging.rst

@@ -61,7 +61,7 @@ argument of :class:`~pytorch_lightning.trainer.trainer.Trainer`)
 
 .. testcode::
 
-    # use only 1% of training data (and use the same training Dataloader (with shuffle off) in val and test)
+    # use only 1% of training data (and use the same training dataloader (with shuffle off) in val and test)
     trainer = Trainer(overfit_batches=0.01)
 
     # or overfit a number of batches

pytorch_lightning/trainer/__init__.py

@@ -688,12 +688,12 @@ def on_train_end(self, trainer, pl_module):
 
 .. warning:: .. deprecated:: 0.8.0.
 
-    Use `overfit_batches`. Will remove 1.0.0.
+    Use `overfit_batches`. Will be removed in 0.10.0.
 
 overfit_batches
 ^^^^^^^^^^^^^^^
-Uses this much data of the training set. If will use the same training set for validation and testing.
-If the training Dataloaders(shuffle=True), Lightning will automatically disable it.
+Uses this much data of the training set. If nonzero, will use the same training set for validation and testing.
+If the training dataloaders have `shuffle=True`, Lightning will automatically disable it.
 
 Useful for quickly debugging or trying to overfit on purpose.
 

pytorch_lightning/trainer/data_loading.py

@@ -265,12 +265,12 @@ def _reset_eval_dataloader(
 
                 # when overfitting, the dataloader should not have sampler
                 if self.overfit_batches > 0:
-                    rank_zero_warn('You requested to overfit but enabled training Dataloader shuffling.'
+                    rank_zero_warn('You requested to overfit but enabled training dataloader shuffling.'
                                    ' We are turning it off for you.')
                     dataloaders[loader_i] = self.replace_sampler(loader, SequentialSampler(loader.dataset))
 
                 else:
-                    rank_zero_warn(f'Your {mode}_dataloader has shuffle=True, it is best practice to turn'
+                    rank_zero_warn(f'Your {mode}_dataloader has `shuffle=True`, it is best practice to turn'
                                    ' this off for validation and test dataloaders.')
 
         if any([dl is None for dl in dataloaders]):
@@ -374,7 +374,7 @@ def determine_data_use_amount(self, overfit_batches: float) -> None:
         if overfit_batches > 0:
             if isinstance(overfit_batches, float) and overfit_batches > 1:
                 raise ValueError('`overfit_batches` when used as a percentage must'
-                                 f' be not 0.0 < x < 1.0 but got {overfit_batches:.3f}.')
+                                 f' be in range 0.0 < x < 1.0 but got {overfit_batches:.3f}.')
             self.limit_train_batches = overfit_batches
             self.limit_val_batches = overfit_batches
             self.limit_test_batches = overfit_batches

pytorch_lightning/trainer/evaluation_loop.py

@@ -22,7 +22,7 @@
 Set how much of the validation set to check
 -------------------------------------------
 
-If you don't want to check 100% of the validation set (for debugging or if it's huge), set this flag
+If you don't want to check 100% of the validation set (for debugging or if it's huge), set this flag.
 
 limit_val_batches will be overwritten by overfit_batches if `overfit_batches > 0`
 
@@ -37,7 +37,7 @@
 Set how much of the test set to check
 -------------------------------------
 
-If you don't want to check 100% of the test set (for debugging or if it's huge), set this flag
+If you don't want to check 100% of the test set (for debugging or if it's huge), set this flag.
 
 limit_test_batches will be overwritten by overfit_batches if `overfit_batches > 0`
 

pytorch_lightning/trainer/trainer.py

@@ -190,12 +190,12 @@ def __init__(
             progress_bar_refresh_rate: How often to refresh progress bar (in steps). Value ``0`` disables progress bar.
                 Ignored when a custom callback is passed to :paramref:`~Trainer.callbacks`.
 
-            overfit_batches: Overfit a percent of training data (float) or a set number of batches (int).
+            overfit_batches: Overfit a percent of training data (float) or a set number of batches (int). Default: 0.0
 
             overfit_pct:
                 .. warning:: .. deprecated:: 0.8.0
 
-                    Use `overfit_batches` instead. Will remove 0.10.0.
+                    Use `overfit_batches` instead. Will be removed in 0.10.0.
 
             track_grad_norm: -1 no tracking. Otherwise tracks that p-norm. May be set to 'inf' infinity-norm.