Updated the docs for PytorchSeq2VecWrapper to specify that mask i…

…s required (#5386)

* Updated docs to have masks rather than lengths

* Changelog

* Rephrased the documentation to better reflect mask

Co-authored-by: Akshita Bhagia <akshita23bhagia@gmail.com>

CHANGELOG.md

@@ -28,6 +28,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - LZMA compression is now supported.
 - Added a way to give JSON blobs as input to dataset readers in the `evaluate` command.
 - Added the argument `sub_module` in `PretrainedTransformerMismatchedEmbedder`
+- Updated the docs for `PytorchSeq2VecWrapper` to specify that `mask` is required rather than sequence lengths for clarity.
 
 ### Changed
 

allennlp/modules/seq2vec_encoders/pytorch_seq2vec_wrapper.py

@@ -35,9 +35,10 @@ class PytorchSeq2VecWrapper(Seq2VecEncoder):
     This is what pytorch's RNN's look like - just make sure your class looks like those, and it
     should work.
 
-    Note that we *require* you to pass sequence lengths when you call this module, to avoid subtle
-    bugs around masking.  If you already have a `PackedSequence` you can pass `None` as the
-    second parameter.
+    Note that we *require* you to pass a binary `mask` of shape
+    (batch_size, sequence_length) when you call this module, to avoid subtle
+    bugs around masking. If you already have a `PackedSequence` you can pass
+    `None` as the second parameter.
     """
 
     def __init__(self, module: torch.nn.modules.RNNBase) -> None: