SageMaker Studio Lab Examples

Example Jupyter notebooks that demonstrate how to build AI/ML learning environment using Amazon SageMaker Studio Lab.

Background

SageMaker Studio Lab is a service for individual data scientist who wants to develop the career toward AI/ML practitioner. You can start your ML journey for free.

This repository introduces you to the way to set up Studio Lab according to your interest area, such as computer vision, natural language processing, etc. And also, we show how to deploy your project to the Amazon SageMaker to become the AI/ML practitioner.

:hammer_and_wrench: Setup

Please follow the Onboard to Amazon SageMaker Studio Lab.

1. Request a Studio Lab account
2. Create a Studio Lab account
3. Sign in to Studio Lab

If you would like to localize the user interface, please follow the instruction for user interface localization.

Usage

1. Read: You can read the notebook in Studio Lab without Studio Lab account. Please feel free to click Open in Studio Lab button in Examples section.
2. Run: You can run the notebook by copying the notebook or git clone the repository to your Studio Lab project.
3. Share: You can share the notebooks through the Git repository such as GitHub. If you add Open in Studio Lab button, the readers can copy the notebook or clone the repository by clicking button.

Examples

Computer Vision

No	Title	Open in Studio Lab
1	Train an image classification model with PyTorch
2	Weather Classification for Disaster Risk Reduction with DenseNet-161

Natural Language Processing

No	Title	Open in Studio Lab
1	Finetune T5 locally for machine translation on COVID-19 Health Service Announcements with Hugging Face

Geospatial Data Science

No	Title	Open in Studio Lab
1	Getting Started With Geospatial Data Analysis
2	Exploratory Analysis for NOAA Weather and Climate Dataset

Generative Deep Learning

No	Title	Open in Studio Lab
1	Introduction to JumpStart - Text to Image
2	Prompting Mistral 7B Instruct

Connect To AWS

No	Title	Open in Studio Lab
1	Using SageMaker Studio Lab with AWS Resources
2	Deploy A Hugging Face Pretrained Model to Amazon SageMaker Serverless Endpoint - Boto3

Custom Environments

We provide .yml files to set up various programming language / framework environments. To use the .yml file, please proceed with the following instruction.

1. Click this button right here –>
2. Click the Copy to Project button
   • Sign-in and Start runtime is needed before it.
3. When prompted, select Clone Entire Repo
4. Click Clone after confirming Open README files. is checked.
   • When No Conda environment file found shown, please Dismiss.
5. Once opening README.md preview, please move to Custom Environments section and click the programming language / specific framework environment link as you need to open .yml file.
6. Right click the opened .yml file tab and select Show in File Browser.
7. Right click the .yml file in the file browser and select Build Conda Environment.
8. Once command completed, please run notebook in the same folder to check the environment. When prompted Select Kearnel, please select the created environment.

Programming language environment

• R environment
• Julia environment

Specific framework environment

• AutoGluon (CPU) environment
  • AutoGluon is AutoML library for quick prototype by state-of-the-art method without expert knowledge.
• fast.ai environment
  • fast.ai is deep learning library which provides state-of-the-art results with high-level API for practitioners and low-level API for expert.
• SciPy environment
  • SciPy is an open-source software for mathematics, science, and engineering.
• Diffusers environment
  • diffusers provides pretrained diffusion models across multiple modalities, such as vision and audio, and serves as a modular toolbox for inference and training of diffusion models.
• RAPIDS environment (external link)
  • RAPIDS provides GPU accelerated data science libraries.
• Geospatial environment
  • Geospatial environment is built from basic libraries for geospatial analysis such as geopandas, shapely, and folium, etc.
• Medical image AI environment
  • Medial image AI environment is built from basic libraries for medical image analysis such as itkwidgets, monai.
• Gradio environment
  • Gradio is an application that is suitable for demonstrating your model through an interactive interface.

Community contents

Here are some more examples from the community.

Studio Lab Examples in GitHub.

Please add amazon-sagemaker-lab tag to your repositories that use Studio Lab! We will pick up the popular repositories in here or our blog.

:balance_scale: License

This project is licensed under the Apache-2.0 License.

:handshake: Contributing

Although we’re extremely excited to receive contributions from the community, we’re still working on the best mechanism to take in examples from external sources. Please bear with us in the short-term if pull requests take longer than expected or are closed.

Please read our contributing guidelines if you’d like to open an issue or submit a pull request.

🔎 References

• SageMaker Studio Lab
• SageMaker Studio Lab document
• Stack Overflow
• Regional examples
  • SageMaker Studio Lab Community Japan