Git commit
https://github.com/leejet/stable-diffusion.cpp/pkgs/container/stable-diffusion.cpp/1178678032?tag=master-cuda
sha256:dcd82f38252a32822dcd0c80672d5948df8e63bb4a3064988e0f7c2bec10c100
Operating System & Version
CachyOS
GGML backends
CUDA
Command-line arguments used
entrypoint: ["/sd-server"] command: [ "--verbose", "--listen-ip", "0.0.0.0", "--diffusion-model", "/models/diffusion_models/flux-2-klein-4b.safetensors", "--llm", "/models/text_encoders/qwen_3_4b.safetensors", "--vae", "/models/vae/flux2-vae.safetensors", "--diffusion-fa", ]
Steps to reproduce
Add this to my compose file:
diffusioncpp:
image: ghcr.io/leejet/stable-diffusion.cpp:master-cuda
entrypoint: ["/sd-server"]
command: [
"--verbose",
"--listen-ip",
"0.0.0.0",
"--diffusion-model",
"/models/diffusion_models/flux-2-klein-4b.safetensors",
"--llm",
"/models/text_encoders/qwen_3_4b.safetensors",
"--vae",
"/models/vae/flux2-vae.safetensors",
"--diffusion-fa",
]
ports:
- 1234:1234
volumes:
# Borrow my ComfyUI models
- $USER_HOME/comfy/ComfyUI/models:/models
- ./sdcpp:/output
restart: unless-stopped
deploy:
resources:
reservations:
devices:
- driver: nvidia
capabilities: [gpu]
Run docker compose up -d
Go to http://localhost:1234 or curl -v http://localhost:1234/sdcpp/v1/capabilities
What you expected to happen
I expect to see a working web ui or a list of model capabilities.
What actually happened
Web UI loads but says "service unavailable". Curl request returns:
* Host localhost:1234 was resolved.
* IPv6: ::1
* IPv4: 127.0.0.1
* Trying [::1]:1234...
* Established connection to localhost (::1 port 1234) from ::1 port 37962
* using HTTP/1.x
> GET /sdcpp/v1/capabilities HTTP/1.1
> Host: localhost:1234
> User-Agent: curl/8.21.0
> Accept: */*
>
* Request completely sent off
< HTTP/1.1 500 Internal Server Error
< Connection: close
< Content-Length: 0
< EXCEPTION_WHAT: filesystem error: status: Operation not permitted [./proc/1/map_files/55c6e8164000-55c6e817d000]
< Access-Control-Allow-Methods: *
< Access-Control-Allow-Credentials: true
< Access-Control-Allow-Headers: *
< Access-Control-Allow-Origin: *
<
* shutting down connection #0
Logs / error messages / stack trace
Curl results provided, no logs show from the server despite the --verbose tag.
Additional context / environment details
No response
Git commit
https://github.com/leejet/stable-diffusion.cpp/pkgs/container/stable-diffusion.cpp/1178678032?tag=master-cuda
sha256:dcd82f38252a32822dcd0c80672d5948df8e63bb4a3064988e0f7c2bec10c100
Operating System & Version
CachyOS
GGML backends
CUDA
Command-line arguments used
Steps to reproduce
Add this to my compose file:
Run
docker compose up -dGo to http://localhost:1234 or
curl -v http://localhost:1234/sdcpp/v1/capabilitiesWhat you expected to happen
I expect to see a working web ui or a list of model capabilities.
What actually happened
Web UI loads but says "service unavailable". Curl request returns:
Logs / error messages / stack trace
Curl results provided, no logs show from the server despite the
--verbosetag.Additional context / environment details
No response