NeuralDialog-LaRL icon indicating copy to clipboard operation
NeuralDialog-LaRL copied to clipboard

Published 20 hours ago • verified publisher icon snakeztc
→

Metadata

PyTorch implementation of latent space reinforcement learning for E2E dialog published at NAACL 2019. It is released by Tiancheng Zhao (Tony) from Dialog Research Center, LTI, CMU

Results 3 NeuralDialog-LaRL issues
Sort by recently updated
recently updated
newest added

How can we get actions before response generation from z?

how can we get the action from z (latent variable) at each dialogue turn? @snakeztc

akashicMarga avatar akashicMarga

Trained Model

1 comment

Would you share your trained model? We have some problems reproducing the same result as in your paper. Thanks!

jincli avatar jincli

Confusion about the diversity metric and other loss terms

Hi, thanks for the nice paper and neat code, well done. After read the paper and check the code, I was confused about several lines in `model_task.py`, from 306 to...

TobiasLee avatar TobiasLee

← Metadata

145
Stars
26
Forks
Watchers

Owner

verified publisher icon snakeztc

Metadata

PyTorch implementation of latent space reinforcement learning for E2E dialog published at NAACL 2019. It is released by Tiancheng Zhao (Tony) from Dialog Research Center, LTI, CMU

Back