No 'context_id' in data
yujianll opened this issue · comments
Hi, I'm trying to run the code. After loading the file, I found the data doesn't contain 'context_id' at this line. Thus throws an error. I'm running the first step, i.e., ython code/joint_model.py -relation_weight 1.0 -entity_weight 0 -data_type "tbd" -batch 4 -model "singletask/pipeline" -epochs 10. I wonder if I ran the script correctly. Thanks!
Agree with yujianll. I looked and the data contained in the pickle files in the data/matres and data/tbd contain lists of dictionaries with the following keys, none of which contain the "context" and "context_id" the EventDataset class expects.
'rel_type', 'rev', 'doc_dictionary', 'event_labels', 'left_event', 'right_event', 'doc_id'
aha - just saw the instructions in the README. We need to run the following data pre-processing scripts first:
featurize_data.py
context_aggregator.py