The tokenizer in llama.cpp probably needs fixing then or it has some other bug. | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		bugglebeetle on June 27, 2024 \| parent \| context \| favorite \| on: Gemma 2: Improving Open Language Models at a Pract... The tokenizer in llama.cpp probably needs fixing then or it has some other bug.

0x7cfe on July 1, 2024 [–]

Definitely. I tried gemma2:27B model with phrases like "translate the following sentence to language X" and it even failed to understand the task and spat out completely irrelevant things, like math formulas.

OTOH, smaller model did it perfectly.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact