zeblok-sdk 1.1.2

Zeblok Python SDK

Zeblok Python SDK

Zeblok-sdk will help integrate your MLOps workflows with the Zeblok Ai-MicroCloud platform.

Components of the Python SDK

Auth

• API Authentication
• Datalake Authentication

Resources

• Plan
• Namespace

Managed Service

• Microservice
• Orchestration-Addons
• Ai-API
• Ai-Pipeline
• DataSet

Interface

Authentication

• API Authentication: APIAuth(app_url, api_access_key,api_access_secret)
• Datalake Authentication: DataSet(api_auth, access_key, secret_key, bucket_name)

Plans

• Get all plans: get_all()
• Get a plan by id: get_by_id(plan_id)
• Validate a plan by id: validate_id(plan_id)
• Get filtered details of a plan: get_filtered_details(plan_id, fields_req)

Namespaces

• Get all namespaces: get_all()
• Get a namespace by id: get_by_id(namespace_id)
• Validate a namespace by id: validate_id(namespace_id)

MicroServices

• Get all microservices: get_all()
• Get a microservice by id: get_by_id(plan_id)
• Validate a microservice by id: validate_id(microservice_id)
• Spawn a microservices by id: spawn(display_name, microservice_id, plan_id, microservice_name, namespace_id, envs, ports, args, command)

Orchestration-Addons

• Get all orchestrations: get_all()
• Get an orchestration by id: get_by_id(orchestration_id)
• Validate an orchestration by id: validate_id(orchestration_id)
• Spawn an orchestrations by id: spawn(orchestration_id, plan_id, namespace_id, orchestration_name, min_workers, max_workers)

AI-APIs

• Get all AI-APIs with deployed state: get_all(state=deployed)
• Get all AI-APIs with ready state: get_all(state=ready)
• Validate an AI-API: validate(image_name, state)
• Create and Spawn an AI-API: create_and_spawn(ai_api_name, model_folder_path, plan_id, namespace_id, ai_api_type)
• Create an AI-API: create(ai_api_name, model_folder_path, ai_api_type)
• Spawn an AI-API: spawn(image_name, namespace_id, plan_id)

AI-Pipelines

• Get all AI-Pipelines with created state: get_all(state=created)
• Get all AI-Pipelines with ready state: get_all(state=ready)
• Validate an AI-Pipeline: validate(image_name, state)
• Create and Spawn an AI-API: create_and_spawn(ai_pipeline_name, ai_pipeline_folder_path, caas_plan_id, ai_pipeline_plan_id, namespace_id)
• Create an AI-Pipeline: create(ai_pipeline_name, ai_pipeline_folder_path, caas_plan_id, ai_pipeline_plan_id, namespace_id)
• Spawn an AI-Pipeline: spawn(ai_pipeline_plan_id, namespace_id, ai_pipeline_image_name)

DataSet

• Check if a bucket exists: bucket_exists()
• Check if an object exists: object_exists(object_name)
• Upload an object: upload_object(local_file_pathname, object_name)
• Download an object: download_object(object_name, local_dir, filename)
• Get object PreSigned URL: get_presigned_url(object_name)

Usage

Checkout the detailed usage of the SDK at Usage.md

Contact Zeblok

Team	Email
Technical Team	[e-mail]
Sales Team	[e-mail]