Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
fixed sentence-length bug when using the tokenizer with sentence pair…
…s. (#60) The `sep_offset_value` in `build_scatter_offsets` method should be computed on just the first sentence of the pair in input otherwise if the second sentence is bigger, the resulting sentence offsets are wrong.
- Loading branch information