tingofurro / summac

Codebase, data and models for the SummaC paper in TACL
https://arxiv.org/abs/2111.09525
Apache License 2.0
75 stars 19 forks source link

Problem when loading the benchmark #6

Open zhehengluoK opened 1 year ago

zhehengluoK commented 1 year ago

hi, I encounter a key error problem using benchmark.py to load the cogensumm dataset. The self.get_cnndm_document func always returns a key error indicating the "aid" is not in the loaded CNN/DM dataset. I can't figure out why this problem happens. Could you please help?

zhehengluoK commented 1 year ago

Hello, my problem is solved when I switched to another device.

Another problem I met when loading the benchmark is the 'cnn_dailymail' datset on huggingface used in benchmark.py seems to have a problem now, so I suggest replacing it with "load_dataset('ccdv/cnn_dailymail')" which is a copy of the original one.