diff options
| author | Georgi Gerganov <ggerganov@gmail.com> | 2023-03-11 01:22:58 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-03-11 01:22:58 +0200 |
| commit | 01eeed8fb1437978603a8523c0b8ea2f6280f5d7 (patch) | |
| tree | 57dcf3218c96aeb7d9493f6409ccf4fef77797ea | |
| parent | 6da2df34ee40301d9ecb126968ec4c0c6195f26d (diff) | |
Update README.md
| -rw-r--r-- | README.md | 2 |
1 files changed, 2 insertions, 0 deletions
@@ -5,6 +5,8 @@ Inference of [Facebook's LLaMA](https://github.com/facebookresearch/llama) model **TEMPORARY NOTICE:** If you observe garbage results, make sure to update to latest master. There was a bug and it was fixed here: https://github.com/ggerganov/llama.cpp/commit/70bc0b8b15b98dca23b28f0c8f5e34b27e424cda +Also, currently the quantized models run **only** on Apple Silicon. On other architectures, you can [use the F16 models](https://github.com/ggerganov/llama.cpp/issues/2#issuecomment-1464615286), but they will be much slower. Support will be [added later](https://github.com/ggerganov/ggml/pull/27) + ## Description The main goal is to run the model using 4-bit quantization on a MacBook. |