match tokenizer and model versions #17

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Open

dcarpintero wants to merge 1 commit into HandsOnLLM:main from dcarpintero:dcarpintero/chapter02-match-tokenizer-model

+1 −1

Contributor

dcarpintero commented Nov 24, 2024

Match tokenizer and model versions:

When looking at the papers, the vocab size of 'deberta-v3' is 128k whereas in 'deberta-base' was 50k, and the internal architectures have also some differences. This might lead to issues.
Therefore it is proposed to update the tokenizer to 'microsoft/deberta-v3-base'


          match tokenizer and model versions

8b6ef85

Contributor

MaartenGr commented Nov 29, 2024

@jalammar Any preference here for updating the tokenizer?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet