IndexError: while accessing the outputs in RBA_celeba.ipynb file

[dileeppanigrahi]

Hi All,
I am currently studying the paper and trying to run this code,
I've trained 'celeb_baseline', 'celeba_uniconf_adv' & 'celeba_domain_independent' models, but while running the RBA_celeba.ipynb file I've faced some issues in the test and dev output indexing ("IndexError: index 40 is out of bounds for axis 1 with size 39") in the below shells. can you please help me to fix this issue?

dev_outputs = dev['output'][:, subclass_idx + [item+39 for item in subclass_idx]]
test_outputs = test['output'][:, subclass_idx + [item+39 for item in subclass_idx]]

IndexError Traceback (most recent call last)
in
----> 1 dev_outputs = dev['output'][:, subclass_idx + [item+39 for item in subclass_idx]]
2 test_outputs = test['output'][:, subclass_idx + [item+39 for item in subclass_idx]]

IndexError: index 40 is out of bounds for axis 1 with size 39

Note :
Shapes of the Dev and Test results :
dev['output'].shape is (19867, 39)
test['output'].shape is (19962, 39) , but in the script we are accessing the 40th index. can you please help me to understand?

[zywangcode]

The RBA method works with domain discriminative training, where the output contains twice the class number.