smaug-bpe

#1
by xxx31dingdong - opened

Why is this model using smaug-bpe instead of llama-bpe? Is this intentional?

Resplendent AI org

It is an issue with llamacpp, you can easily fix using an inbuilt script to change the pre-tokenizer. Or you could just download my model quants from mradermacher which work out of the box.

Sign up or log in to comment