Chris Thompson
Chris Thompson
I have. I suppose my only hesitation is that PGXN seems obscure even though its the only extension repository I can see. Is it in prevalent use?
There's a [PR](https://github.com/ggerganov/llama.cpp/pull/4070) from the deepseek team about this. Basically, their tokenizer needs to be supported in llama.cpp for this to work.
Do two things: `pip install sentencepiece -U` and use convert-hf-to-gguf.py This worked for me, and the GGUF output was able to perform inference.