Update sparse.py

Author: soumith

torch/nn/modules/sparse.py

@@ -18,7 +18,7 @@ class Embedding(Module):
  max_norm (float, optional): If given, will renormalize the embeddings to always have a norm lesser than this
  norm_type (float, optional): The p of the p-norm to compute for the max_norm option
  scale_grad_by_freq (boolean, optional): if given, this will scale gradients by the frequency of
- the words in the dictionary.
+ the words in the mini-batch.
 
  Attributes:
  weight (Tensor): the learnable weights of the module of shape (num_embeddings, embedding_dim)