Error when converting huihui-ai/Llama-3.2-3B-Instruct-abliterated: Received parameters not in model: lm_head.weight.

#36
by Felladrin - opened
MLX Community org

Error when trying to convert https://huggingface.co./huihui-ai/Llama-3.2-3B-Instruct-abliterated:

image.png

The same error happens when trying to convert https://huggingface.co./chuanli11/Llama-3.2-3B-Instruct-uncensored

image.png

The error happens with any quantization selected.

MLX Community org

make sure you're logged in by clicking the button at the top of page first.

MLX Community org

Ah, I'm logged in:

image.png

MLX Community org

The config has tier_word_embeddings = True so it's a bit odd that it's trying to load an LM head weight. I wonder if it's a bug in the config?

MLX Community org

Sign up or log in to comment