You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi Andevani,
It's me again.
I use my corpus to make prepro.py build json and h5 files.
But this happens
Inserting the special UNK token
Assigned 10000 to val, 10000 to test.
Traceback (most recent call last):
File "prepro_utf8.py", line 240, in
Main(params)
File "prepro_utf8.py", line 173, in main
L, label_start_ix, label_end_ix, label_length = encode_captions(imgs, params, wtoi)
File "prepro_utf8.py", line 150, in encode_captions
Assert np.all(label_length > 0), 'error: some caption had no words?'
AssertionError: error: some caption had no words?
Can you help me with my doubts?
Thank you very much.
The text was updated successfully, but these errors were encountered:
Hi Andevani,
It's me again.
I use my corpus to make prepro.py build json and h5 files.
But this happens
Inserting the special UNK token
Assigned 10000 to val, 10000 to test.
Traceback (most recent call last):
File "prepro_utf8.py", line 240, in
Main(params)
File "prepro_utf8.py", line 173, in main
L, label_start_ix, label_end_ix, label_length = encode_captions(imgs, params, wtoi)
File "prepro_utf8.py", line 150, in encode_captions
Assert np.all(label_length > 0), 'error: some caption had no words?'
AssertionError: error: some caption had no words?
Can you help me with my doubts?
Thank you very much.
The text was updated successfully, but these errors were encountered: