Marigold
A package manager and runtime for local AI applications. One self-hosted model layer serves every application on the machine: applications install by name from registries, share one model cache and one GPU queue, and reach models through a typed API with OpenAI-compatible chat and embeddings.
Underneath, typed operations -- a capability class, a model and an input set -- produce immutable results that applications compose.
Marigold runs as three separate layers on one host:
- the platform -- a model cache, a catalogue, a worker and an API, shared by everything on the host;
- applications -- a list of required models plus the code that uses them, installed into the cache by name;
- instances -- an application's code running in its own container, reaching the platform only through the API.
Applications hold no weights. Many instances share one model cache and one worker, which is what lets a fleet of agents run on a single local GPU and the same application run unchanged on larger hardware.
The documentation is in docs/, and published at marigold.run/docs.
Status
Early stage. The interfaces an application is written against --
marigold.toml and models.yaml, the instance contract, the API
routes and the registry index (protocol: 1) -- are stable; see
What is stable. The CLI and the
configuration format may change. Issues and questions are welcome; for a pull request, open an
issue first to discuss the change before writing code.
Prerequisites
- Docker and Docker Compose
- NVIDIA Container Toolkit, for a GPU worker
- Python 3.12, for the
marigoldcommand - A Hugging Face token, for gated models
Quickstart
pip install bayis-marigold
marigold cache init
marigold registry add https://examples.marigold.run/
marigold application install chat
marigold cache populate chat
marigold application start chat
registry add records the examples registry. install fetches the
chat application from it. cache populate downloads the models it
declares; this is the long step. start brings the platform up and
runs the application: Open WebUI at http://localhost:3000, backed by
a local instruct model. The walkthrough is in
Getting started.
Documentation
| Getting started | install, first application, OpenAI-compatible client |
| Applications | what an application is; building, publishing, installing, finding |
| Examples, Showcase | the project's registries |
| Porting | bringing an existing AI application to Marigold |
| Configuration | marigold-host.toml: host settings, registries, storage, resource limits |
| Deployment | from a laptop to your own cloud account |
| Model types | every type, route, input and output; adding a type |
| API | submit and poll, binary outputs, OpenAI-compatible endpoints |
| Architecture, Principles | how the system is divided, and why |
Repository structure
package/src/
api/ FastAPI application; routes/ is a package
models/ model handlers, one file per model type
shared/ enums, registry, database, outputs, usage
cli/ the `marigold` command
main.py parser
commands.py command implementations
config.py system and application TOML, with provenance
paths.py the cache layout
archives.py writing and extracting application archives
applications.py identity, installation, resolution
compose.py compose projects, environment, invocation
compose/ compose files and the multi-stage Dockerfile
tools/
cache_cli.py the cache container's entrypoint
model_cache_shared.py cache providers
Example applications live in marigold-examples, published as the registry at examples.marigold.run.
Metadata
Release files for bayis-marigold 0.8.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| bayis_marigold-0.8.2.tar.gz | 172.0 kB | Without provenance | Details for bayis_marigold-0.8.2.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| bayis_marigold-0.8.2-py3-none-any.whl | Python 3 | none | any | Details for bayis_marigold-0.8.2-py3-none-any.whl |
Total release size: 350.9 kB
Release files / bayis_marigold-0.8.2.tar.gz
| Download URL | bayis_marigold-0.8.2.tar.gz |
|---|---|
| Size | 172.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3b5db0ada6fc8cb34c7b33029f9e1f3e2e82315f0777e0c04f3915ad9048190b
|
|
BLAKE2b-256 checksum How to use checksums |
2b3c0a33c80aaf07000cab6eba46bed5725837e221f3ad3d719ce9310f086776
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.16
|
Release files / bayis_marigold-0.8.2-py3-none-any.whl
| Download URL | bayis_marigold-0.8.2-py3-none-any.whl |
|---|---|
| Size | 178.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
49bd82cea2ef66065b605403ffd38351511eb1a4595c2118a81979a02a2b82ba
|
|
BLAKE2b-256 checksum How to use checksums |
fadc926d2fe7e43d0717da077e82998786115251ab388c0f8506cbb611bc4c44
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.16
|
Security for bayis-marigold 0.8.2
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.