Failed to load model

#1
by QuamtumX - opened

I've downloaded "Meta-Llama-3-8B-Instruct-abliterated-v3.Q6_K.gguf"

What is causing this message:

"llama.cpp error: 'error loading model vocabulary: unknown pre-tokenizer type: 'smaug-bpe''

I'm using LM Studio, and preset "Llama 3"

I have the same error perhaps QuantFactory has shipped the wrong tokenizer as the original
quants at https://huggingface.co/failspy/Meta-Llama-3-8B-Instruct-abliterated-v3-GGUF/ do seem to work

Thanks for your response Ictrev, I'm going to download that one instead.

Update: it works fine.

Sign up or log in to comment