baalajimaestro/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Update README.md

Browse source
This commit is contained in:
Matt Hoffner 2023-06-10 15:59:26 -07:00 committed by GitHub
parent 74fbaae157
commit 4eb245afd8
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 6 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
README.md
View file

@ -41,7 +41,7 @@ bash Miniforge3-MacOSX-arm64.sh
``` ```
Otherwise, while installing it will build the llama.ccp x86 version which will be 10x slower on Apple Silicon (M1) Mac. Otherwise, while installing it will build the llama.ccp x86 version which will be 10x slower on Apple Silicon (M1) Mac.
### Installation with OpenBLAS / cuBLAS / CLBlast ### Installation with OpenBLAS / cuBLAS / CLBlast / Metal
`llama.cpp` supports multiple BLAS backends for faster processing. `llama.cpp` supports multiple BLAS backends for faster processing.
Use the `FORCE_CMAKE=1` environment variable to force the use of `cmake` and install the pip package for the desired BLAS backend. Use the `FORCE_CMAKE=1` environment variable to force the use of `cmake` and install the pip package for the desired BLAS backend.
@ -64,6 +64,11 @@ To install with CLBlast, set the `LLAMA_CLBLAST=1` environment variable before i
CMAKE_ARGS="-DLLAMA_CLBLAST=on" FORCE_CMAKE=1 pip install llama-cpp-python CMAKE_ARGS="-DLLAMA_CLBLAST=on" FORCE_CMAKE=1 pip install llama-cpp-python
``` ```
To install with Metal (MPS), set the `LLAMA_METAL=on` environment variable before installing:
```bash
CMAKE_ARGS="-DLLAMA_METAL=on" FORCE_CMAKE=1 pip install llama-cpp-python
```
## High-level API ## High-level API