BentoML is an open platform that simplifies ML model deployment and enables you to serve your models at production scale in minutes
👉 Pop into our Slack community! We're happy to help with any issue you face or even just to meet you and hear what you're working on :)
This repository is a collection of machine learning projects that utilizes BentoML for model serving. The goal is to demonstrate real-world BentoML usage and best practices in productionizing a machine learning model for serving.
Note: You are looking at gallery examples for BentoML 1.0 version, which is still under early beta release. For prior stable versions (0.13.x), see the 0.13-LTS branch.
BentoML 1.0 preview release is required for running gallery projects here:
pip install bentoml --pre- Scikit-learn Iris Classifier: https://github.com/bentoml/gallery/tree/main/quickstart
- PyTorch MNIST: https://github.com/bentoml/gallery/tree/main/pytorch_mnist
Each gallery project is under its own folder, typically containing the following files:
| file name | description |
|---|---|
| README.md | a step-by-step guide running the project Python scripts from CLI |
| {PROJECT_NAME}.ipynb | a jupyter notebook shows the same workflow but from notebook environment |
| requirements.txt | required PyPI packages for this project |
| train.py | a python script for training an ML model and saving it with BentoML |
| import_model.py | import an existing trained model to BentoML |
| service.py | python code that defines the bentoml.Service instance for serving |
| bentofile.yaml | the bento build file for building the service into a Bento |
| .bentoignore | files to exclude from build directory, when building a Bento |
| benchmark.py | a python script that tests the baseline performance of the final model server created |
If you have issues running these projects or have suggestions for improvement, use Github Issues 🐱
If you are interested in contributing new projects to this repo, let's talk 🥰 - Join us on Slack and share your idea in #bentoml-dev channel
Before you create a Pull Request, make sure:
- Follow the basic structures and naming conventions of other existing gallery projects
- Ensure your project runs with the latest version of BentoML