Extended dataset

[xiaoqian-shen]

Great work! I am curious about whether you will release the extended dataset with referential text.

[snarkyidiot]

I would also appreciate the release of this dataset

[xiaoqian-shen]

Btw, could you provide the pretrained character classifier for convenient evaluation?

[trahman8]

We re-used existing datasets. In the dataloader, we modify the ref texts. See ldm/data/flintstones_data.py, ldm/data/mugen_data.py and ldm/data/proro_data.py to generate ref text.

[xiaoqian-shen]

Thanks for your reply. This issue then could be closed.

[xiaoqian-shen]

Hi! I am trying to evaluate the accuracy of generated stories. Could you provide the pretrained weight of character and background classifier for convenient evaluation?