alex-berard / seq2seq

Attention-based sequence to sequence learning
Apache License 2.0
388 stars 122 forks source link

reinforcement learning detail? #3

Closed zhangzibin closed 7 years ago

zhangzibin commented 7 years ago

Hi, Thanks for you good work. I found you code support reinforcement learning. Can you give more detail? What paper you implement?

choltz95 commented 7 years ago

I am also interested in some notes on this - the idea is clear, but I cannot find a paper.

alex-berard commented 7 years ago

The algorithm which was implemented is called REINFORCE. Its application to sequence to sequence models is described here: https://arxiv.org/abs/1511.06732