| # Copyright 2023 Mistral AI and The HuggingFace Inc. team. All rights reserved. | |
| # | |
| # Licensed under the Apache License, Version 2.0 (the "License"); | |
| # you may not use this file except in compliance with the License. | |
| # You may obtain a copy of the License at | |
| # | |
| # http://www.apache.org/licenses/LICENSE-2.0 | |
| # | |
| # Unless required by applicable law or agreed to in writing, software | |
| # distributed under the License is distributed on an "AS IS" BASIS, |
Good question! I am collecting human data on how quantization affects outputs. See here for more information: ggml-org/llama.cpp#5962
In the meantime, use the largest that fully fits in your GPU. If you can comfortably fit Q4_K_S, try using a model with more parameters.
See the wiki upstream: https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix
This gist is out of date and I can no longer help much, as I got rid of my Mac.
Please visit T2 Linux website for more and better information:
This gist is just a compilation of the hard work that others have put in. I'm not a software developer, so if there are any mistakes or better ways of doing things, I'd appreciate any suggestions. Here's a list of the real heroes who made this possible:
