Downloading the model
python download_model.py
Building container
docker build -t stablediffusion .
Running
docker run -p 4000:80 -v /Users/brianfoody/apps/personal/storytime/packages/avatars2/images:/images stablediffusion docker run -v /Users/brianfoody/apps/personal/storytime/packages/avatars2/images:/images stablediffusion docker run -v /home/ec2-user/stablediffusionaws/images:/images stablediffusion
Running on AWS Ec2
git clone https://github.com/brianfoody/stablediffusionaws.git
python3 -m pip install --upgrade diffusers transformers accelerate python3 -m pip install torch torchvision torchaudio python3 download_model.py
Connecting to ec2 and Pushing to ecs
See SD_AWS.md
Grabbing files for testing
scp -i ./SensiveStableDiffusionUsWest1.pem ./img.png ec2-user@ip-172-31-9-213:/home/ec2-user/stablediffusionaws/images/output.png
Checking the drivers (cuda and CuDNN)
nvcc --version
cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
or
nvidia-smi
FROM nvidia/cuda:11.7-cudnn8-runtime-ubuntu18.04