Yulin Chen

> Thanks for pointing this out. Did you find any actual inconsistencies? Yes. For topic 1, some inconsistency examples are (the format is `file (source, target) relation1 relation2`) ``` 1_18ecbplus.xml...

what do you mean by "结果不稳定"("unstable prediction"), can you provide further context of the specific code you have run?

Hi, currently the `__init__` function in `PrefixTuningTemplate` cannot handle `OPT`, you may need to modify it by yourself. Pull request is welcomed. https://github.com/thunlp/OpenPrompt/blob/8756a8cf5af4e7e01db3d612d74cb4cd1d927aba/openprompt/prompts/prefix_tuning_template.py#L64

Doing NER with prompt learning is not a straightforward task, because in NER you are basically doing classification for every token in the sentence. You can refer to this paper...

I am not sure I understand your question correctly. It seems that you are not using {"mask"} and therefore not using `verbalizer` on the whole. That way you probably should...

A straightforward way is to concat your examples into one string and make it part of the template. Your template will look like: ``` ... {"placeholder": "text_a"} [some kind of...

Hi, you may try upgrading you python version to python 3.7 and above.

Hi, thanks for your question. Yes, supporting local deployment on low-resource device is definitely a must-do future work. We are currently focusing on training and releasing larger and better models,...

Hi, thanks for your interest in our work. The reason is that we saved the delta weights in torch.float32, so it is almost twice as large.

Yes, we put everything under the role "user" for openai chatgpt api. So for example, if you already have the first round of conversation as below ``` User: user utterance...