Removes the group_add configuration for video and render groups. These permissions are often redundant or handled by the container runtime, and removing them simplifies the configuration and prevents potential GID mismatch issues.
Add a new ComfyUI service to the Docker Compose configuration, specifically tailored for AMD GPUs using ROCm.
- Uses the rocm/pytorch base image for hardware acceleration
- Includes an initialization script to clone and install ComfyUI on first boot
- Configures necessary hardware device mappings (/dev/dri, /dev/kfd) and environment variables for AMD compatibility
- Sets up persistent storage and network integration with the existing proxy and AI internal networks
Introduce a new compose.yml to deploy Ollama with AMD GPU hardware acceleration on Docker Swarm.
- Uses the ROCm-enabled Ollama image.
- Mounts necessary GPU devices (/dev/kfd and /dev/dri) with root permissions for hardware access.
- Sets HSA_OVERRIDE_GFX_VERSION to 10.3.0 to support Navi 21 (6900 XT) cards.
- Configures persistent storage on local SSD and integrates with an external proxy network.
- Includes deployment constraints to target nodes labeled with GPUs.
