GMFTBY

Lol, this is also the motivation of this repo. But the transformer based model seems more powerful than the RNN based models. If you have some ideas, we can make...

Hi, i compare the GPT-2 model (transformers), and use the data to train it from scratch. GPT-2 model can achieve better distinct score (better diversity). But the BLEU and embedding-based...

And also in `tsd_net.py` ```python def supervised_loss(self, pz_proba, pm_dec_proba, z_input, m_input): pz_proba, pm_dec_proba = pz_proba[:, :, :cfg.vocab_size].contiguous(), pm_dec_proba[:, :, :cfg.vocab_size].contiguous() pr_loss = self.pr_loss(pz_proba.view(-1, pz_proba.size(2)), z_input.view(-1)) m_loss = self.dec_loss(pm_dec_proba.view(-1, pm_dec_proba.size(2)), m_input.view(-1))...

先创建目录 ~/.cache/oh-my-earth

Sorry, I didn't show the tutorial for this repo. Please run the `utils.py` to generate the model-ready data ```bash python utils.py --mode dataset --dataset xxxx ``` **Dataset's name should be...

As for the datasets, you can check the following benchmarks: * Xiaohuangji: https://github.com/fate233/dgk_lost_conv * Tencent: arXiv 2018 / A Manually Annotated Chinese Corpus for Non-task-oriented Dialogue * Dailydialog: https://github.com/Sanghoon94/DailyDialogue-Parser *...