An AI Engineer Guide to Comet Model Registry Platform

Explore the Comet Model Registry Platform in this comprehensive 32-minute tutorial for AI engineers. Learn how to monitor, log, register, version, and deploy machine learning models using Comet's sophisticated system. Follow along as the instructor demonstrates key features, including experiment initialization, Keras ML model training, model monitoring, and exporting models to the Comet server. Discover how to register models in the Comet Model Registry, generate new models with updated configurations, and download models from the registry. Gain practical insights into real-time model monitoring, risk and drift assessment, and efficient model management throughout the machine learning lifecycle.

- Video Start
- Video Content Intro
- Comet Monitoring Platform overview
- Notebook intro at Google Colab
- Comet ML Pipeline Monitoring Tutorial Intro
- Comet Model Export API Documentation
- Tutorial Starts now
- Comet Experiment Initialization
- Keras ML Experiment started
- Deep Learning Model Training Started
- Model Monitoring at Comet started
- Saving Keras Model to Disk
- Comet Desktop at Notebook
- Exporting Model to Comet Server
- Registering Model to Comet Model Registry
- Generating new Model with new Config
- Adding updated model with new version
- Downloading Model from Comet model registry
- Completing Experiment
- Recap
- Credits