yorko / bert-finetuning-catalyst Goto Github PK

Hey Yorko!

Thank you for the library and the youtube explanations, ran into it while exploring BERT!

Apologize in advance for a beginner question.

I'm trying to run your code on my data. However, it seems that I cannot specify anything but 5 clusters, otherwise the code crashes with

File "/home/svet/gitprojects/bert-finetuning-catalyst/venv/lib/python3.7/site-packages/torch/nn/functional.py", line 2265, in nll_loss
ret = torch._C._nn.nll_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index)
IndexError: Target 4 is out of bounds.

(I have 4 clusters)

May be I missed something in your explanations but your sdg_id column has > 5 different values, this also confused me.

Would be really grateful if you responed.

Cheers,

Hi,

first of all, thanks for these wonderful repository and tutorial. I have an isse, though. When i run the script i get this error code: 'Indexing with integers (to access backend Encoding for a given batch index) is not available when using Python based tokenizers'

Do you know why it is happening?

To recreate the issue, i did the following:

cloned the repository
python src/train.py

Thanks