search

harvardnlp / seq2seq-attn

Sequence-to-sequence model with LSTM encoder/decoders and attention
http://nlp.seas.harvard.edu/code
MIT License
1.26k stars 278 forks source link

beam.lua needs a -model argument #11

Closed srush closed 8 years ago

srush commented 8 years ago

also relatedly not giving -outfile should default to stdout