[DataLoader] __getitems__ added to description of Dataset API and b…

…etter supported within `Subset` (pytorch#100375)

DataLoader supports batched loading from Mapped Datasets.

This is the fetcher's implementation of auto-detection of batch loading support.

torch.utils.data._utils.fetch._MapDatasetFetcher
```
class _MapDatasetFetcher(_BaseDatasetFetcher):
    def fetch(self, possibly_batched_index):
        if self.auto_collation:
            if hasattr(self.dataset, "__getitems__") and self.dataset.__getitems__:
                data = self.dataset.__getitems__(possibly_batched_index)
            else:
                data = [self.dataset[idx] for idx in possibly_batched_index]
```

Description of Dataset API now shows this feature.

Additionally, Subset dataset now supports `__getitems__` if parent dataset supports it.
Pull Request resolved: pytorch#100375
Approved by: https://github.com/ejguan, https://github.com/NivekT

torch/utils/data/dataset.py

@@ -41,7 +41,10 @@ class Dataset(Generic[T_co]):
     data sample for a given key. Subclasses could also optionally overwrite
     :meth:`__len__`, which is expected to return the size of the dataset by many
     :class:`~torch.utils.data.Sampler` implementations and the default options
-    of :class:`~torch.utils.data.DataLoader`.
+    of :class:`~torch.utils.data.DataLoader`. Subclasses could also
+    optionally implement :meth:`__getitems__`, for speedup batched samples
+    loading. This method accepts list of indices of samples of batch and returns
+    list of samples.
 
     .. note::
       :class:`~torch.utils.data.DataLoader` by default constructs a index
@@ -52,6 +55,10 @@ class Dataset(Generic[T_co]):
     def __getitem__(self, index) -> T_co:
         raise NotImplementedError("Subclasses of Dataset should implement __getitem__.")
 
+    # def __getitems__(self, indices: List) -> List[T_co]:
+    # Not implemented to prevent false-positives in fetcher check in
+    # torch.utils.data._utils.fetch._MapDatasetFetcher
+
     def __add__(self, other: 'Dataset[T_co]') -> 'ConcatDataset[T_co]':
         return ConcatDataset([self, other])
 
@@ -296,6 +303,14 @@ def __getitem__(self, idx):
             return self.dataset[[self.indices[i] for i in idx]]
         return self.dataset[self.indices[idx]]
 
+    def __getitems__(self, indices: List[int]) -> List[T_co]:
+        # add batched sampling support when parent dataset supports it.
+        # see torch.utils.data._utils.fetch._MapDatasetFetcher
+        if callable(getattr(self.dataset, "__getitems__", None)):
+            return self.dataset.__getitems__([self.indices[idx] for idx in indices])  # type: ignore[attr-defined]
+        else:
+            return [self.dataset[self.indices[idx]] for idx in indices]
+
     def __len__(self):
         return len(self.indices)