Skip to content

AWS SageMaker Course Release v10

Choose a tag to compare

@github-actions github-actions released this 25 Jul 15:38
· 17 commits to main since this release

馃彔 AWS SageMaker Course Materials

This release includes both data files and source code for the AWS SageMaker course.

馃搳 Data Files (california-housing-data.tar.gz)

  • california_housing.csv - Original dataset (20,640 rows)
  • california_housing_train.csv - Training set (16,512 rows)
  • california_housing_test.csv - Test set (4,128 rows)

馃 Pre-trained Model

  • trained_model.joblib - Ready-to-use LinearRegression model (Simple scikit-learn model)

馃悕 Source Code (aws-sagemaker-course-source.tar.gz)

  • train.py - SageMaker training script for California Housing dataset
  • sagemakerTraining.py - Creates SageMaker training jobs (Estimator + ModelTrainer)
  • sagemakerDeployServerless.py - Model deployment examples (3 serverless endpoints)
  • sagemakerDeployRealTime.py - Model deployment examples (3 serverless endpoints + 1 real-time endpoint)
  • template_requirements.txt - Python package dependencies to run templates

馃殌 Quick Start

# Download everything
wget https://github.com/CodeSignal-Learn/course_building-aws-sagemaker/releases/download/v10/california-housing-data.tar.gz
wget https://github.com/CodeSignal-Learn/course_building-aws-sagemaker/releases/download/v10/aws-sagemaker-course-source.tar.gz
wget https://github.com/CodeSignal-Learn/course_building-aws-sagemaker/releases/download/v10/trained_model.joblib

# Extract files
tar -xzf california-housing-data.tar.gz
tar -xzf aws-sagemaker-course-source.tar.gz

# Install dependencies and run
pip install -r template_requirements.txt
python sagemakerTraining.py          # Training jobs
python sagemakerDeployServerless.py  # Models deployment
python sagemakerDeployRealTime.py    # Models deployment

馃搱 Dataset Features

  • Original: MedInc, HouseAge, AveRooms, AveBedrms, Population, AveOccup, Latitude, Longitude, MedHouseVal
  • Train/Test: Additional RoomsPerHousehold feature + capped values at 95th percentile
  • Split: 80% train / 20% test (random_state=42)

Generated automatically from commit 079103adea636ae1a8247dca5f8adafc12e27bff