bos_token is not added in tokenizing text by default

#5
by khaimai - opened

I found that tokenizer of microsoft/Phi-3-medium-4k-instruct didn't add the bos_token by default like that from microsoft/Phi-3-mini-4k-instruct
Screen Shot 2024-05-24 at 11.28.49.png

So I assume that bos_token should be added to the beginning of the tokens ?

Sign up or log in to comment