Reduce the Amount of Code Needed to Build ML Models
summary
Shows how to turn code from a Jupyter Notebook (exploratory analysis, data cleaning, feature engineering, training, tuning and evaluation) into a repeatable, scalable ML pipeline on AWS by adding only a few lines of code. Demonstrates using AWS services and orchestration tools to package and automate the end-to-end machine learning workflow.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 4:42Import libraries and define parameters
- 6:57Function to prepare the dataset
- 10:20Automatic container creation
- 12:36Running the function as a Processing job
- 16:49Function for model training
- 20:25Running the function as a Training job
- 22:44Function for performance evaluation
- 28:07Class to deploy the model
- 32:02Deploying the model as an Endpoint
- 34:50Invoking the Endpoint
- 38:30Parameters and outputs of functions
- 49:36Running the Pipeline
related talks
25mDesigning an AWS Well Architected Framework for AI/ML
32mAutomating Production Production Level Machine Learning on AWS
17mBetter availability and automated code reviews with ML
44mPragmatic Machine Learning in the Cloud
31mProductizing Machine Learning over Big Data with AWS tools
41mLeveraging open-source and 3rd party tools for ML on AWS
41mBuilding Serverless Machine Learning Models in the Cloud
34mDeploying a Machine Learning System That Only Runs Locally