bodywork-pipeline-with-aporia-monitoring
bodywork
bodywork-pipeline-with-aporia-monitoring | bodywork | |
---|---|---|
1 | 8 | |
4 | 430 | |
- | - | |
0.0 | 0.0 | |
almost 2 years ago | 9 months ago | |
Jupyter Notebook | Python | |
MIT License | GNU Affero General Public License v3.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
bodywork-pipeline-with-aporia-monitoring
-
Calling Aporia from Bodywork Pipelines to Monitor Models in Production
Monitoring models for drift and degradation is not easy - theoretically or practically. In this example project we show to outsource these problems to Aporia’s model monitoring platform, by using their Python client from within a Bodywork pipeline.
bodywork
- Deployment automation for ML projects of all shapes and sizes
-
A tutorial on how to handle prediction uncertainty in production systems, by using Bayesian inference and probabilistic programs
how to deploy it to Kuberentes using Bodywork.
-
[P] [D] How are you approaching prediction uncertainty in ML systems?
I usually turn to generative models - e.g. probabilistic programs and Bayesian inference. I’ve written-up my thoughts on how to engineer these into a ‘production system’ deployed to Kubernetes, using PyMC and Bodywork (an open-source ML deployment tool that I contribute to).
- Bodywork: MLOps tool for deploying ML projects to Kubernetes
-
Tool for mapping executable Python modules to Kubernetes deployments
I’m one of the core contributors to Bodywork, an open-source tool for deploying machine learning projects developed in Python, to Kubernetes.
-
[P] [D] The benefits of training the simplest model you can think of and deploying it to production, as soon as you can.
I’ve had many successes with this approach. With this in mind, I’ve put together an example of how to make this Agile approach to developing machine learning systems a reality, by demonstrating that it takes under 15 minutes to deploy a Scikit-Learn model, using FastAPI with Bodywork (an open-source MLOps tool that I have built).
- bodywork - MLOps for Python and K8S
- bodywork-ml/bodywork-core - MLOps automation for Python and Kubernetes
What are some alternatives?
evidently - Evaluate and monitor ML models from validation to production. Join our Discord: https://discord.com/invite/xZjKRaNp8b
NuPIC - Numenta Platform for Intelligent Computing is an implementation of Hierarchical Temporal Memory (HTM), a theory of intelligence based strictly on the neuroscience of the neocortex.
VevestaX - 2 Lines of code to track ML experiments + EDA + check into Github
gensim - Topic Modelling for Humans
bodywork-pymc3-project - Serving Uncertainty with Bayesian inference, using PyMC3 with Bodywork
PaddlePaddle - PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice (『飞桨』核心框架,深度学习&机器学习高性能单机、分布式训练和跨平台部署)
ML-Workspace - 🛠 All-in-one web-based IDE specialized for machine learning and data science.
Crab - Crab is a flexible, fast recommender engine for Python that integrates classic information filtering recommendation algorithms in the world of scientific Python packages (numpy, scipy, matplotlib).
ml-pipeline-engineering - Best practices for engineering ML pipelines.
TFLearn - Deep learning library featuring a higher-level API for TensorFlow.
MLOps - End to End toy example of MLOps
PyBrain