Shun Zheng comments

Results 16 comments of


                                            Shun Zheng

question about your ChFi-nAnn dataset and bert model

We create the ChFinAnn dataset by distant supervision, and processing details are included by the main paper (Section 4) and the supplementary material (Section A.1). I will refactor the BERT...

question about your ChFi-nAnn dataset and bert model

Sounds cool! Currently, Doc2EDAG assumes that the input is a sequence of sentences, so it is suitable for event-related documents, which are mainly expressed by the natural language. As for...

question about your ChFi-nAnn dataset and bert model

@pjfeng What you have mentioned is a very challenging topic, and many startups also worked on it in recent years. Discussions are welcome. Thanks for your interests in our work.

question about your ChFi-nAnn dataset and bert model

@xiaocuigit Founder is about extracting inter-entity relations from richly formatted documents, while Doc2EDAG focuses on extracting various event records (each with multiple entities) from a text document. At present, Doc2EDAG...

activation_checkpointing error when using --fsdp

@vtien Genius! Using the REENTRANT checkpointing option also works for me. As described in https://pytorch.org/docs/stable/checkpoint.html, these two types of checkpointing methods do have certain differences. I will try to see...

activation_checkpointing error when using --fsdp

@irenedea I have tried to use 'attn_implementation=eager', still triggering a smilar error. Interestingly, using the REENTRANT checkpointing strategy does not help me for this case. The bug is really weird.