{"id":21844004,"url":"https://github.com/altescy/metamaker","last_synced_at":"2025-07-08T17:35:53.723Z","repository":{"id":39761314,"uuid":"424939602","full_name":"altescy/metamaker","owner":"altescy","description":"⚗️ Simple command line tool to train and deploy your machine learning models with AWS SageMaker","archived":false,"fork":false,"pushed_at":"2023-02-08T02:17:16.000Z","size":459,"stargazers_count":5,"open_issues_count":7,"forks_count":1,"subscribers_count":2,"default_branch":"main","last_synced_at":"2025-06-18T07:51:45.002Z","etag":null,"topics":["fastapi","machine-learning","poetry","python","sagemaker"],"latest_commit_sha":null,"homepage":"https://pypi.org/project/metamaker/","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/altescy.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2021-11-05T12:25:29.000Z","updated_at":"2024-01-23T15:48:11.000Z","dependencies_parsed_at":"2025-04-14T12:20:43.144Z","dependency_job_id":null,"html_url":"https://github.com/altescy/metamaker","commit_stats":null,"previous_names":[],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/altescy/metamaker","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/altescy%2Fmetamaker","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/altescy%2Fmetamaker/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/altescy%2Fmetamaker/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/altescy%2Fmetamaker/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/altescy","download_url":"https://codeload.github.com/altescy/metamaker/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/altescy%2Fmetamaker/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":264315081,"owners_count":23589704,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["fastapi","machine-learning","poetry","python","sagemaker"],"created_at":"2024-11-27T22:18:06.388Z","updated_at":"2025-07-08T17:35:53.705Z","avatar_url":"https://github.com/altescy.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"metamaker\n=========\n\n[![Actions Status](https://github.com/altescy/metamaker/workflows/CI/badge.svg)](https://github.com/altescy/metamaker/actions/workflows/main.yaml)\n[![License](https://img.shields.io/github/license/altescy/metamaker)](https://github.com/altescy/metamaker/blob/master/LICENSE)\n[![Python version](https://img.shields.io/pypi/pyversions/metamaker)](https://github.com/altescy/metamaker)\n[![pypi version](https://img.shields.io/pypi/v/metamaker)](https://pypi.org/project/metamaker/)\n\nSimple command line tool to train and deploy your machine learning models with AWS SageMaker\n\n## Features\n\nmetamaker enables you to:\n\n- Build a docker image for training and inference with [poetry](https://python-poetry.org/) and [FastAPI](https://fastapi.tiangolo.com/)\n- Train your own machine learning model with SageMaker\n- Deploy inference endpoint with SageMaker\n\n## Usage\n\n1. Create poetry project and install metamaker\n\n```\n❯ poetry new your_module\n❯ cd your_module\n❯ poetry add metamaker\n```\n\n2. Define scripts for traning and inference in `main.py`\n\n```main.py\nfrom pathlib import Path\nfrom typing import Any, Dict\n\nfrom metamaker import MetaMaker\n\n# Import your model, and input/output data classs:\n#\n#   Model  ... machine learning model class you want to use\n#   Input  ... input data class for inference\n#   Output ... ouput data class for inference\n#\n# Note that the Input and Output are used as type hints to\n# create API endpoint with FastAPI like below:\n#\n#   @fastapi_app.post(\"/invocations\")\n#   def predict(data: Input) -\u003e Output:\n#       ...\nfrom your_module import Model, Input, Output\n\napp = MetaMaker[Model, Input, Output]()\n\n@app.trainer\ndef train(\n    dataset_path: Path,\n    artifact_path: Path,\n    hyperparameters: Dict[str, Any],\n) -\u003e None:\n    model = Model(**hyperparameters)\n    model.train(dataset_path / \"train.csv\")\n    model.save(artifact_path / \"model.tar.gz\")\n\n@app.loader\ndef load(artifact_path: Path) -\u003e Model:\n    return Model.load(artifact_path / \"model.tar.gz\")\n\n@app.predictor\ndef predict(model: Model, data: Input) -\u003e Output:\n    return model.predict(data)\n```\n\n3. Write metamaker configs in `metamaker.yaml`\n\n```metamaker.yaml\n# Specify metamaker handler like: `path.to.module:app_name`\nhandler: main:app\n\n# dataset_path and artifact_path should be directories and end with '/'\ndataset_path: s3://your-bucket/path/to/dataset/\nartifact_path: s3://your-bucket/path/to/artifacts/\n\nhyperparameter_path: ./hparams.yaml\n\nimage:\n  name: metamaker\n  includes:\n    - your_module/\n    - main.py\n  excludes:\n    - __pycache__/\n    - '*.py[cod]'\n\ntraining:\n  execution_role: arn:aws:iam::xxxxxxxxxxxx:role/SageMakerExecutionRole\n  instance:\n    type: ml.m5.large\n    count: 1\n\ninference:\n  endpoint_name: your_endpoint\n  instance:\n    type: ml.t2.meduim\n    count: 1\n```\n\n4. Build docker image and push to ECR\n\n```\nmetamaker build --deploy .\n```\n\n5. Train your model with SageMaker and deploy endpoint\n\n```\nmetamaker sagemaker train --deploy\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faltescy%2Fmetamaker","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faltescy%2Fmetamaker","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faltescy%2Fmetamaker/lists"}