For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Python module
max.pipelines.lib.artifact
Exports a pipeline to, and loads it from, a .m model artifact.
An artifact is a directory holding metadata.json (provenance and file
checksums), artifact.json (one section per pipeline component), and any
files the components reference, such as the compiled MEF.
Classes
ArtifactComponent | A pipeline component that can be stored in a model artifact. |
|---|---|
ArtifactFile | A file stored inside the .m artifact directory. |
ArtifactMetadata | Where an artifact came from and the checksums of its files. |
ArtifactType | A value a component may return from __serialize__. |
HuggingFaceRef | A reference to a Hugging Face repo, or a file in one, at a revision. |
ModelArtifact | Serializable serve-able pipeline. |
Exceptions
ArtifactError | Raised when an artifact cannot be written, read, or validated. |
|---|
Functions
deserialize | Deserializes a model artifact from a .m directory. |
|---|---|
serialize | Serializes a model artifact to a .m directory. |