Skip to main content

:::note Generated reference Generated from honeyframeapi 0.1.8 — do not edit by hand. :::

honeyframeapi.model

Saved-model handle — the ML Lab registry (versioned scikit-learn models).

models = project.list_models()                 # [SavedModel, ...]
m = project.get_model(7)                       # by model_id
m.metrics                                      # {"accuracy": 0.91, ...}
m.activate()                                   # make this the active version
m.delete()

Models are CREATED by train_model / cluster_model recipes on the flow canvas (or via project.create_recipe(recipe_type="train_model", ...) then recipe.run()); this wrapper reads/manages the registry over /api/models: metrics, feature importance, versions, activation.

class SavedModel​

SavedModel.get_info()​

get_info() -> dict

Full report row: metrics, feature_importance, versions[].

SavedModel.name​

property

name: str

SavedModel.version​

property

version: int

SavedModel.task_type​

property

task_type: str

SavedModel.metrics​

property

metrics: dict

SavedModel.feature_importance​

property

feature_importance: dict

SavedModel.feature_columns​

property

feature_columns: list

SavedModel.list_versions()​

list_versions() -> list[dict]

All versions of this model name, newest first.

SavedModel.activate()​

activate() -> dict

Make this version the active one within its name (score-by-name recipes pick it up).

SavedModel.delete()​

delete() -> None

Delete this version (registry row + artifact dir). The active version with siblings refuses with 409 — activate another first.

Verified against honeyframeapi 0.1.8.