Support no pre-fetching (#11606)

Author: carmocca

CHANGELOG.md

@@ -72,6 +72,9 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added `LightningModule.lr_scheduler_step` ([#10249](https://github.com/PyTorchLightning/pytorch-lightning/pull/10249))
 
 
+- Added support for no pre-fetching to `DataFetcher` ([#11606](https://github.com/PyTorchLightning/pytorch-lightning/pull/11606))
+
+
 - Added `opt_idx` to scheduler config if not assigned by user ([#11247](https://github.com/PyTorchLightning/pytorch-lightning/pull/11247))
 
 

pytorch_lightning/utilities/fetching.py

@@ -201,19 +201,15 @@ def _no_op_batch_to_device(batch: Any) -> Any:
 
 
 class DataFetcher(AbstractDataFetcher):
-
-    """This class is used to control batch fetching flow. By default, the ``fetching_function`` will pre-fetch a
-    batch in advance to detect the end of the iteration.
+    """This class is used to control batch fetching flow.
 
     Args:
-        prefetch_batches: Number of batches to be pre-fetched. Lightning will pre-fetch
-            at least 1 batch for tracking the latest batch.
+        prefetch_batches: Number of batches to pre-fetch. Pre-fetching at least 1 batch is necessary to properly track
+            whether a batch is the last one (available with :attr:`self.done`).
         store_on_device: Whether to store the pre-fetched batches on device.
     """
 
     def __init__(self, prefetch_batches: int = 1, store_on_device: bool = True) -> None:
-        if prefetch_batches < 1:
-            raise MisconfigurationException("`prefetch_batches` should at least be 1.")
         super().__init__(prefetch_batches=prefetch_batches)
         self.store_on_device = store_on_device
         self.batch_to_device: Callable[[Any], Any] = _no_op_batch_to_device
@@ -240,19 +236,31 @@ def prefetching(self) -> None:
                 break
 
     def fetching_function(self) -> Tuple[Any, bool]:
+        assert self.dataloader_iter is not None
         if self.batches:
+            # there are pre-fetched batches already from a previous `prefetching` call.
+            # consume one
             batch = self.batches.pop(0)
-        else:
-            # empty iterator, no prefetching done
-            raise StopIteration
-        if not self.done:
-            assert self.dataloader_iter is not None
             try:
+                # refill the consumed batch
                 self._fetch_next_batch(self.dataloader_iter)
             except StopIteration:
+                # no more batches to fetch. we are done only if all pre-fetched batches were returned
+                self.done = not self.batches
+        elif not self.done:
+            # this will run only when no pre-fetching was done.
+            try:
+                self._fetch_next_batch(self.dataloader_iter)
+                # consume the batch we just fetched
+                batch = self.batches.pop(0)
+            except StopIteration as e:
                 self.done = True
+                raise e
+        else:
+            # the iterator is empty
+            raise StopIteration
         self.wait()
-        return self.move_to_device(batch), len(self.batches) == 0
+        return self.move_to_device(batch), self.done
 
     def _fetch_next_batch(self, iterator: Iterator) -> None:
         start_output = self.on_fetch_start()

tests/utilities/test_fetching.py

@@ -40,28 +40,32 @@ def __iter__(self):
             yield 2
             yield 3
 
-    for prefetch_batches in range(1, 5):
+    for prefetch_batches in range(5):
+        iterator = DataFetcher(prefetch_batches=prefetch_batches)
+        assert iterator.prefetch_batches == prefetch_batches
+
         if use_combined_loader:
             loader = CombinedLoader([DataLoader(IterDataset()), DataLoader(IterDataset())])
-            expected = [
-                ([tensor([1]), tensor([1])], False),
-                ([tensor([2]), tensor([2])], False),
-                ([tensor([3]), tensor([3])], True),
-            ]
         else:
             loader = DataLoader(IterDataset())
-            expected = [(1, False), (2, False), (3, True)]
-        iterator = DataFetcher(prefetch_batches=prefetch_batches)
-        assert iterator.prefetch_batches == prefetch_batches
         iterator.setup(loader)
 
         def generate():
-            generated = []
-            for idx, data in enumerate(iterator, prefetch_batches + 1):
-                assert iterator.fetched == 3 if iterator.done else idx
-                generated.append(data)
+            generated = [(iterator.fetched, *data) for i, data in enumerate(iterator, prefetch_batches + 1)]
+            assert iterator.fetched == 3
+            assert iterator.done
             return generated
 
+        is_last_batch = [False, False, prefetch_batches > 0]
+        fetched = list(range(prefetch_batches + 1, 4))
+        fetched += [3] * (3 - len(fetched))
+        if use_combined_loader:
+            batches = [[tensor(1), tensor(1)], [tensor(2), tensor(2)], [tensor(3), tensor(3)]]
+        else:
+            batches = [1, 2, 3]
+        expected = list(zip(fetched, batches, is_last_batch))
+        assert len(expected) == 3
+
         assert generate() == expected
         # validate reset works properly.
         assert generate() == expected
@@ -71,9 +75,9 @@ class EmptyIterDataset(IterableDataset):
         def __iter__(self):
             return iter([])
 
-    dataloader = DataLoader(EmptyIterDataset())
+    loader = DataLoader(EmptyIterDataset())
     iterator = DataFetcher()
-    iterator.setup(dataloader)
+    iterator.setup(loader)
     assert not list(iterator)