Merged Documentation

Documentation for model_deployment_worker

Functionality

Handles the model deployment process in the inference worker. It receives a task identifier and calls the deployment utility to initiate model deployment procedures.

Parameters

• task_id: A unique string identifier for the deployment task.

Usage

• Purpose - Triggers model deployment workflows by invoking the underlying deployment handler.

Example

from embedding_studio.workers.inference.worker import \
    model_deployment_worker

task_id = "example_deployment_task_id"
model_deployment_worker(task_id)

---

Documentation for model_deletion_worker

Functionality

The model_deletion_worker is a Dramatiq actor that handles deletion of an embedding model from the Triton Inference Server. It validates the deletion request, checks plugin support, and safely removes model directories.

Parameters

• task_id: A string identifier for the deletion task to process.

Usage

• Purpose: Processes deletion tasks by delegating to the handle_deletion function, ensuring validations and safe removals.

Example

When a deletion task is enqueued, call:

model_deletion_worker("task_identifier")