WizModel

What Is WizModel?

WizModel is a platform that simplifies deploying machine learning models to production. It provides a single, unified API for ML inference, so you can get predictions from any model without writing the complex code typically required for deployment and inference. It also includes a demo UI for testing models without extra coding.

Who’s Behind WizModel?

WizModel was developed by the team behind Cog2, a tool for packaging machine learning models into production-ready containers. The company offers this unified API for machine learning inference, streamlining the process of getting models into production.

Key Features

• Unified API: A single API for ML inference, regardless of the model.
• Production-Ready Deployment: Get models deployed and ready for production use.
• Simplified Inference: Make predictions without writing extensive code or managing dependencies.
• Interactive Demo UI: Test models with a built-in interface.
• Cloud Deployment: Supports deploying models to the cloud.
• Automated Packaging: Automatically packages models into production-ready containers.
• Local & Cloud Predictions: Run predictions on your own machine or in the cloud.
• REST API Access: Access inference through a REST API.

Who Should Use WizModel?

WizModel is aimed at data scientists, machine learning engineers, AI researchers, and AI/ML developers.

Getting Started with WizModel

1. Set Up Your Environment: Create a new project and add your configurations in the cog.yaml file.
2. Write Your Prediction Script: Create a predict.py script that defines how predictions run on your model.
3. Test Locally and Push: Run predictions locally with the cog2 predict command, then push your model to the server with cog2 push.
4. Run in the Cloud: Call the REST API — for example from Python — using the provided URL and your API key.

Make sure your configurations, Python packages, and prediction logic are set up correctly, as detailed in the user guide.