An AI experiment for non-experts

Nvidia AI Workbench is a platform with all the tools you need to develop, train, and deploy artificial intelligence (AI) models, leveraging the power of Nvidia GPUs.

All-in-one: Integrated tools for AI and machine learning.

GPU optimization: Leverage Nvidia GPUs for fast computation.

Collaboration: Easily share models.

Scalability: Deploy to cloud or on-premises infrastructure.

Prerequisites: Basic knowledge of Python and machine learning.

Installation: Install Docker, Nvidia Docker, download AI Workbench, and set up your environment.

Code editor: Develop with Jupyter Notebook.

Dataset management: Preprocess and load data.

Monitoring: Monitor model performance in real time.

Training: Define and train models on GPUs with TensorFlow or PyTorch.

Deployment: Export and run your model on cloud or on-premises.

If you don't have a powerful GPU, AI Workbench offers simulated modes and access to cloud GPUs.

Start with simple models.

Check out the documentation and join the community for support.

To simulate using the GPU without any available hardware, you can set up a CPU-only version of the container:

docker run -p 8888:8888 -v /path/to/your/code:/workspace -it nvidia/ai-workbench:cpu