Confuse about the SDP dataset

Alibaba-NLP / ACE

[ACL-IJCNLP 2021] Automated Concatenation of Embeddings for Structured Prediction

Other

296 stars 44 forks source link

Confuse about the SDP dataset #39

Closed iamownt closed 2 years ago

iamownt commented 2 years ago

When I run train.py with config-psd_parsing_model.yaml, I encounter “FileNotFoundError: [Errno 2] No such file or directory: '/home/rnx/.flair/datasets/enhanced_ud/PSD' Error, which indicates that neither do I have train, Dev and test of PSD, DM and PAS datasets nor I know about how to preprocess them. Could you provide some tips on how to get and preprocess data sets? Do I need to buy dataset from https://catalog.ldc.upenn.edu/LDC2016T10？

wangxinyu0922 commented 2 years ago

You should buy the dataset from LDC and use the preprocess script at https://github.com/tdozat/Parser-v3/tree/master/scripts to convert the file into conll format

iamownt commented 2 years ago

Thanks for your help. Wish you great successes in your future PhD carrer.