Repository navigation
c9n 0.8.11 — Self-hosting installer
c9n 0.8.11
Speicher & Modelle auf der GPU-Worker-Seite
- Neue Übersicht „Speicher & Modelle“ je Worker: wie viel Speicher der Rechner insgesamt hat, wie viel davon die GPU
nutzen darf, und jedes Modell als Balken im selben Maßstab – mit seinen Teilen (z. B. Videomodell: Text-Encoder,
Generator, Decoder). - Man sieht, welches Modell gerade geladen ist und welches nur nacheinander in den Speicher passt. Es ist immer nur ein
Modell geladen; der Worker wechselt vor jedem Auftrag. - Die Größen meldet der Worker selbst (neue
worker.py).
Self-hosted software for guided pilots. Linux amd64 / Docker Compose.
Application image archive and matching installer. The source commit, exact image ID and SHA-256 checksums are recorded in release.json. No application source is published here.
Verify the matching SHA256SUMS file before installation. Installed systems offer this release under Settings -> Hosting; the owner starts the update there.
Local use does not require the optional community gateway. GPU sharing and credits are not yet available.
Product and pilot enquiries: https://c9n.app/