Alternate quantizations.

#6
by ZeroWw - opened

https://huggingface.co/ZeroWw/Llama-3-8B-Instruct-Gradient-4194k-GGUF

These are my own quantizations (updated almost daily).

The difference with normal quantizations is that I quantize the output and embed tensors to f16.
and the other tensors to 15_k,q6_k or q8_0.
This creates models that are little or not degraded at all and have a smaller size.
They run at about 3-6 t/sec on CPU only using llama.cpp
And obviously faster on computers with potent GPUs

You can find all my quantizations here: https://huggingface.co/RobertSinclair

fast

Sign up or log in to comment