Skip to content

Latest commit

 

History

History
63 lines (62 loc) · 2.36 KB

README.md

File metadata and controls

63 lines (62 loc) · 2.36 KB

💬 Stable Diffusion Radeon ROCm 📷

AUTOMATIC1111 implemented using Gradio library.

sudo usermod -a -G render,video $LOGNAME
wget https://raw.githubusercontent.com/hqnicolas/StableDiffusionROCm/v1.9.3-RX7XT/AMD-ROCm-Drivers/prepare.sh
sudo chmod 777 prepare.sh
sudo ./prepare.sh
  • For RDNA 3 cards RX7000:
export HSA_OVERRIDE_GFX_VERSION=11.0.0
  • install Docker
wget https://raw.githubusercontent.com/hqnicolas/StableDiffusionROCm/v1.9.3-RX7XT/docker.sh
sudo chmod 777 docker.sh
sudo ./docker.sh
  • install Docker Compose
sudo apt-get update
sudo apt-get install docker-compose-plugin -y
sudo apt-get install docker-compose -y
  • Install Stable Diffusion ROCm
git clone --branch v1.9.3-RX7XT https://github.com/hqnicolas/StableDiffusionROCm.git
cd StableDiffusionROCm
sudo docker-compose build stablediff-rocm
sudo docker compose up -d stablediff-rocm
  • CTRL+C to stop UP Compose
  • Copy Models to StableDiffusionROCm/stablediff-models
cp v1-5-pruned.ckpt StableDiffusionROCm/stablediff-models
  • Start Stable Diffusion ROCm
sudo docker start -a stablediff-rocm-runner
  • Stop Stable Diffusion ROCm
sudo docker stop stablediff-rocm-runner

To see a prompt from your GPU usage.

watch /opt/rocm-6.0.2/bin/rocm-smi
sudo docker ps
sudo docker exec -it stablediff-rocm-runner /bin/bash -c "watch rocm-smi"

To check python version.

sudo docker exec -it stablediff-rocm-runner /bin/bash -c "python -V"