Implicit Unlikelihood Training: Improving Neural Text Generation with Reinforcement Learning.

Evgeny Lagutin Daniil Gavrilov Pavel Kalaidin

Published in: CoRR (2021)

Keyphrases

text generation
reinforcement learning
network architecture
natural language generation
neural network
fitted q iteration
training set
state space
supervised learning
recurrent networks
training phase
learning algorithm
natural language
training algorithm
test set
training examples
multi agent