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

Update README.md

Browse source
This commit is contained in:
John 2023-06-19 15:38:22 +02:00 • committed by GitHub
parent 1f421dddde
commit 911226625a
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 29 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

29
README.md
View file

@ -22,9 +22,36 @@ _Important: The Falcon 7B model features tensor sizes that do not support K-type
* On linux Q5_1 7B user reports a batch token ingestion context memory issue, with -b 1 it's gone. Not reproduced on Windows * On linux Q5_1 7B user reports a batch token ingestion context memory issue, with -b 1 it's gone. Not reproduced on Windows
* VRAM scratch/overhead calculation on CUDA can fail - if GPU RAM fills to 100% manually reduce the layers of --ngl until it fits * VRAM scratch/overhead calculation on CUDA can fail - if GPU RAM fills to 100% manually reduce the layers of --ngl until it fits
**How to compile:**
```
How to build:
1) Recommended with cmake: (change the CUBLAS flag to 0 to disable CUDA requirements and support)
git clone
cd ggllm.cpp
rm -rf build; mkdir build; cd build
cmake -DLLAMA_CUBLAS=1 ..
cmake --build . --config Release
# find binaries in ./bin
**CUDA (cuda-integration branch, not in master yet):** 2) Installing on WSL (Windows Subsystem for Linux)
# I am getting slightly better timings on WSL than native windows, though currently mmap does not appear to work in WSL (--no-mmap) - either a binary difference or something else
#Choose a current distro:
wsl.exe --list --online
wsl --install -d distro
# cmake 3.16 is required and the cuda toolset
# If you run an old distro you can upgrade (like apt update; apt upgrade; apt full-upgrade; pico /etc/apt/sources.list/; apt update; apt upgrade; apt full-upgrade; apt autoremove; lsb_release -a); then wsl --shutdown and restart it
# install cuda WSL toolkit
wget https://developer.download.nvidia.com/compute/cuda/repos/wsl-ubuntu/x86_64/cuda-keyring_1.0-1_all.deb
dpkg -i cuda-keyring_1.0-1_all.deb
apt-get update; apt-get -y install cuda
# you might need to add it to your path:
export LD_LIBRARY_PATH="/usr/local/cuda-12.1/lib64:$LD_LIBRARY_PATH"
export PATH="/usr/local/cuda-12.1/bin:$PATH"
# now start with a fresh cmake and all should work
```
**CUDA:**
Only some tensors supported currently, only mul_mat operation supported currently Only some tensors supported currently, only mul_mat operation supported currently
q3_k timing on 3090 of Falcon 40B: q3_k timing on 3090 of Falcon 40B:
falcon_print_timings: prompt eval time = 702.55 ms / 3 tokens ( 234.18 ms per token) falcon_print_timings: prompt eval time = 702.55 ms / 3 tokens ( 234.18 ms per token)