getting a ConfigAttrtibuteError when running Infer speech from audio in Collab

[garphillips]

I've followed all the steps up until that point, created a key in hugging face, added a youtube link, given it name, specified all settings in "Infer speech from audio" and then on runtime I get the following error :

`---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
in
237 #return tiny2large, large2tiny, whispers_tokens
238
--> 239 token_large_index_segmentations = whisper_transmit_meta_across_alignment(
240 whispers,
241 large2tiny,

/usr/local/lib/python3.8/dist-packages/vktrs/asr.py in whisper_transmit_meta_across_alignment(whispers, large2tiny, whispers_tokens)
93 rec_large = {'token':whispers_tokens['large'][i]}
94 for j in result:
---> 95 rec_tiny = token_tinyindex_segmentations[j]
96 if not rec_large.get('start'):
97 rec_large['start'] = rec_tiny['start']

KeyError: 16`

Can someone tell me what I'm doing wrong?

[garphillips]

@dmarx do you have any recommendations?