Issues with the current code

[ashim95]

Hi @lushleaf ,

Thank you for releasing the code.

• While running the code for imdb, I get the following error:

  Generate perturbation set                                                                                                                                    
  Traceback (most recent call last):                                                                                                                             
  File "data_process.py", line 375, in <module>                                                                                                                  
  main()                                                                                                                                                     
  File "data_process.py", line 372, in main                                                                                                                      get_perturbation_set(dataset, data_path, similarity_threshold = similarity_threshold, perturbation_constraint = perturbation_constraint)                   File "data_process.py", line 279, in get_perturbation_set                                                                                                      tem_embd_matrix[_, :] = vocab[tem_key_list[_]]['vec']  
  File "env_safer/lib/python3.6/site-packages/networkx/classes/reportviews.py", line 187, in __getitem__               
  return self._nodes[n]                                                                                                                                      
  File "env_safer/lib/python3.6/site-packages/networkx/classes/coreviews.py", line 284, in __getitem__             
  raise KeyError(f"Key {key} not found")                                                                                                                   
  KeyError: 'Key 0 not found' 

Can you help me fix the error?

• I am also trying to run your code on other datasets like stanford sentiment dataset sst-2 and MNLI. It seems the codebase requires some files like *_vocab.pkl and _vocab_pca.pkl etc. Can you please let me know how to generate these files? Or, In general how to run the code on a new dataset?

Thank you,

[Peeaak]

Hi, i faced this problem, too . Luckily, i found the bug. the type of the tem_key_list is not list. it couldn't to traversal. you should make it to be a list.