Solution for errors while running BERT loss, T5 in Google Collab

Course Q&A Natural Language Processing NLP with Attention Models
1

Hi All,

In the BERT loss ungraded lab [C4_W3_Colab_BERT_Loss_Model.ipynb], if you encounter a ModuleNotFoundError: No module named ‘t5’ error while running the statement,

vocab_size = trax.data.vocab_size(vocab_type = ‘sentencepiece’, vocab_file = ‘sentencepiece.model’, vocab_dir = path + “/models/”)

doing a !pip -q install t5 prior would solve the issue.

Similarly, a !pip -q install t5 is needed before running import T5 in cell 2 of ungraded lab T5 [C4_W3_Colab_T5_SQuAD_Model.ipynb]

Thank you.

2

Thank you @vgumpina, for this. We’ll add the line of code in the colabs. Didn’t realise they were missing.