Merge pull request #9593 from gakada/tcmalloc
Try using TCMalloc on Linux by default
This commit is contained in:
commit
1ffb44b0b2
|
@ -120,6 +120,7 @@ sudo pacman -S wget git python3
|
||||||
bash <(wget -qO- https://raw.githubusercontent.com/AUTOMATIC1111/stable-diffusion-webui/master/webui.sh)
|
bash <(wget -qO- https://raw.githubusercontent.com/AUTOMATIC1111/stable-diffusion-webui/master/webui.sh)
|
||||||
```
|
```
|
||||||
3. Run `webui.sh`.
|
3. Run `webui.sh`.
|
||||||
|
4. Check `webui-user.sh` for options.
|
||||||
### Installation on Apple Silicon
|
### Installation on Apple Silicon
|
||||||
|
|
||||||
Find the instructions [here](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Installation-on-Apple-Silicon).
|
Find the instructions [here](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Installation-on-Apple-Silicon).
|
||||||
|
|
|
@ -43,4 +43,7 @@
|
||||||
# Uncomment to enable accelerated launch
|
# Uncomment to enable accelerated launch
|
||||||
#export ACCELERATE="True"
|
#export ACCELERATE="True"
|
||||||
|
|
||||||
|
# Uncomment to disable TCMalloc
|
||||||
|
#export NO_TCMALLOC="True"
|
||||||
|
|
||||||
###########################################
|
###########################################
|
||||||
|
|
21
webui.sh
21
webui.sh
|
@ -113,13 +113,13 @@ case "$gpu_info" in
|
||||||
printf "Experimental support for Renoir: make sure to have at least 4GB of VRAM and 10GB of RAM or enable cpu mode: --use-cpu all --no-half"
|
printf "Experimental support for Renoir: make sure to have at least 4GB of VRAM and 10GB of RAM or enable cpu mode: --use-cpu all --no-half"
|
||||||
printf "\n%s\n" "${delimiter}"
|
printf "\n%s\n" "${delimiter}"
|
||||||
;;
|
;;
|
||||||
*)
|
*)
|
||||||
;;
|
;;
|
||||||
esac
|
esac
|
||||||
if echo "$gpu_info" | grep -q "AMD" && [[ -z "${TORCH_COMMAND}" ]]
|
if echo "$gpu_info" | grep -q "AMD" && [[ -z "${TORCH_COMMAND}" ]]
|
||||||
then
|
then
|
||||||
export TORCH_COMMAND="pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/rocm5.2"
|
export TORCH_COMMAND="pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/rocm5.2"
|
||||||
fi
|
fi
|
||||||
|
|
||||||
for preq in "${GIT}" "${python_cmd}"
|
for preq in "${GIT}" "${python_cmd}"
|
||||||
do
|
do
|
||||||
|
@ -172,15 +172,30 @@ else
|
||||||
exit 1
|
exit 1
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
# Try using TCMalloc on Linux
|
||||||
|
prepare_tcmalloc() {
|
||||||
|
if [[ "${OSTYPE}" == "linux"* ]] && [[ -z "${NO_TCMALLOC}" ]] && [[ -z "${LD_PRELOAD}" ]]; then
|
||||||
|
TCMALLOC="$(ldconfig -p | grep -Po "libtcmalloc.so.\d" | head -n 1)"
|
||||||
|
if [[ ! -z "${TCMALLOC}" ]]; then
|
||||||
|
echo "Using TCMalloc: ${TCMALLOC}"
|
||||||
|
export LD_PRELOAD="${TCMALLOC}"
|
||||||
|
else
|
||||||
|
printf "\e[1m\e[31mCannot locate TCMalloc (improves CPU memory usage)\e[0m\n"
|
||||||
|
fi
|
||||||
|
fi
|
||||||
|
}
|
||||||
|
|
||||||
if [[ ! -z "${ACCELERATE}" ]] && [ ${ACCELERATE}="True" ] && [ -x "$(command -v accelerate)" ]
|
if [[ ! -z "${ACCELERATE}" ]] && [ ${ACCELERATE}="True" ] && [ -x "$(command -v accelerate)" ]
|
||||||
then
|
then
|
||||||
printf "\n%s\n" "${delimiter}"
|
printf "\n%s\n" "${delimiter}"
|
||||||
printf "Accelerating launch.py..."
|
printf "Accelerating launch.py..."
|
||||||
printf "\n%s\n" "${delimiter}"
|
printf "\n%s\n" "${delimiter}"
|
||||||
|
prepare_tcmalloc
|
||||||
exec accelerate launch --num_cpu_threads_per_process=6 "${LAUNCH_SCRIPT}" "$@"
|
exec accelerate launch --num_cpu_threads_per_process=6 "${LAUNCH_SCRIPT}" "$@"
|
||||||
else
|
else
|
||||||
printf "\n%s\n" "${delimiter}"
|
printf "\n%s\n" "${delimiter}"
|
||||||
printf "Launching launch.py..."
|
printf "Launching launch.py..."
|
||||||
printf "\n%s\n" "${delimiter}"
|
printf "\n%s\n" "${delimiter}"
|
||||||
|
prepare_tcmalloc
|
||||||
exec "${python_cmd}" "${LAUNCH_SCRIPT}" "$@"
|
exec "${python_cmd}" "${LAUNCH_SCRIPT}" "$@"
|
||||||
fi
|
fi
|
||||||
|
|
Loading…
Reference in New Issue