Giters

abhimishra91 / transformers-tutorials

Github repo with tutorials to fine tune transformers for diff NLP tasks

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Classifier ouput features are 4 in Model, then why this printed as 1?

shivanandmn opened this issue · comments

commented

transformers-tutorials/transformers_multiclass_classification.ipynb

Line 527 in 9f12834

" (l3): Linear(in_features=768, out_features=1, bias=True)\n",

commented

Hi @shivanandmn, thank you for pointing that out. The prompt output is from my previous model that had only 1 neuron in the final layer.

That should be 4 for the current network. However, there is no change in the model performance for this particular dataset. I will re-run the notebook and update it with the correct prompts. Rest assured you can take the notebook as it is and execute.