QingyueWang

About data

1

In your paper, you said that ". The raw conversational data were collected via Amazon Mechanical Turk, with annotations provided by domain experts. In total, we have labeled 280 dialogues...

config报错

1

你好，我从提示位置下载了Albert的预训练模型，并尝试直接运行文件/albert/albert_total.py，得到报错 AttributeError: 'BertConfig' object has no attribute 'ln_type'。看起来问题是是缺少In_type参数。我查看了albert模型的config文件也没有这个In_type参数。请问如何解决呢？

In your paper, the rewarder network is modeled a simple feed-forward neural network. When I try to understand it thought this code, I found that it was modeled a LSTM....

About the results of this model

10

Hi, I rerun this project with the same experimental setting. But the joint accuracy only achieves around 55% on multiwoz2.1 dataset. Is there any special version of the transformers or...

Did you do any experiments on WOZ2.0 dataset on STAR model?

2

Did you do any experiments on WOZ2.0 dataset on STAR model? The STAR model is a strong baseline on MultiWOZ dataset. But I evaluated it on WOZ dataset(a single domain...

I was reading through trigger_optimization.py and had a quick question about line 566. I noticed that the optimization seems to be carried out without using a validation set — is...