agiresearch/OpenP5

KeyError when generating CDs collaborative indexing method dataset

liyang2019 opened this issue · 2 comments

Hi,

I got the following error when generating the CDs collaborative indexing method dataset using

sh generate_dataset_1.sh 

Error:
Traceback (most recent call last):
File "/usr/local/google/home/lyliyang/OpenP5/./src/src_llama/generate_dataset.py", line 133, in
main(args)
File "/usr/local/google/home/lyliyang/OpenP5/./src/src_llama/generate_dataset.py", line 36, in main
reindex_user_seq_dict, item_map = indexing.collaborative_indexing(args.data_path, args.dataset, user_sequence_dict,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/google/home/lyliyang/OpenP5/src/src_llama/utils/indexing.py", line 145, in collaborative_indexing
reindex_user_sequence_dict = reindex(user_sequence_dict, user_map, item_map)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/google/home/lyliyang/OpenP5/src/src_llama/utils/indexing.py", line 319, in reindex
reindex_user_sequence_dict[uid] = [item_map[i] for i in items]

Hello, could you provide more details? I tried running the code with provided data but didn't encounter the same error.
image

It looks like it maybe due to some data corruption issue, I re-downloaded the CDs data, and now there is no error now.

Thanks!