Alignments generated by force_align is different from the original alignment.

Hi,

I followed the steps in issue 33 and saved fwd_params, fwd_err, rev_params and rev_err after running fast_align on the training corpus. Later I compared the alignment results by running fast_align from scratch and running force_align.py with the saved params on that training corpus.

I was expecting fast_align produces exactly the same alignments either from scratch or by force_align, but I got different alignments.

I'm sure I followed the correct steps to run these comparisons, so I'm wondering why the alignments are differnt.

Reproduce the comparison described in issue 33 using fast_align from scratch and force_align.py with the saved fwd_params, fwd_err, rev_params, and rev_err on the training corpus. Trace why the two runs produce different alignments, then document or correct the behavior so the expected relationship is clear and verified by the comparison.

Written by the indexing model from the issue text.