what is the difference between "actor-critic training" and "REINFORCE-critic" in the README.md ?

rizar / actor-critic-public

The source code for "An Actor Critic Algorithm for Structured Prediction"

MIT License

167 stars 43 forks source link

Closed haoyusoong closed 6 years ago

haoyusoong commented 6 years ago

haoyusoong commented 6 years ago

I may have figured it out. "2 .actor-critic training" is the step3 of algorithm 2 in the paper .