scalvin1

You are missing some python package (protobuf?). Pull it in with pip install. What you describe is what I gathered from https://voorloopnul.com/blog/quantize-and-run-the-original-llama3-8b-with-llama-cpp/ and I managed to get a working gguf...

Your comments are beside the point. What I want is clear instructions to convert the model **downloaded directly from Meta** to the best-possible feature complete Q8_0 gguf. I want to...

As final words in this neglected issue, I summarize my findings. To the commenter above: huggingface is not the official repository. You download it from Zuckerboy's facebook here https://llama.meta.com/llama-downloads/ What...

There is not much to add anymore, it all seems to be working the way it was described. Closing issue.