Managing Models Cloudera AI allows data scientists to build, deploy, and manage models as REST APIs to serve predictions. Models overviewModel Training and Deployment OverviewThis section provides an overview of model training and deployment using Cloudera AI.Challenges with Machine Learning in productionOne of the hardest parts of Machine Learning (ML) is deploying and operating ML models in production applications. These challenges fall mainly into the following categories: model deployment and serving, model monitoring, and model governance.Using Cloudera AI RegistryCloudera AI Registry is the core enabler for MLOps, or DevOps for machine learning.Creating and deploying a modelUsing Cloudera AI, you can create any function within a script and deploy it to a REST API. In a Cloudera AI project, this is typically a predict function that accepts an input and returns a prediction based on the model's parameters.Usage guidelines for deploying models with Cloudera AIConsider these guidelines when deploying models with Cloudera AI.Known Issues and Limitations with Model Builds and Deployed ModelsModel Request and Response FormatsEvery model function in Cloudera AI takes a single argument in the form of a JSON-encoded object, and returns another JSON-encoded object as output. This format ensures compatibility with any application accessing the model using the API, and gives you the flexibility to define how JSON data types map to your model's datatypes. Testing calls to a ModelSecuring ModelsYou can secure your Cloudera AI models using Access keys or API keys.Workflows for active ModelsThis topic walks you through some nuances between the different workflows available for re-deploying and re-building models. Technical metrics for ModelsYou can observe the operation of your models by using charts provided for technical metrics. These charts can help you determine if your models are under- or over-resourced, or are experiencing some problem.Debugging issues with ModelsThis topic describes some common issues to watch out for during different stages of the model build and deployment process.Deleting a ModelConfiguring model request payload sizeModel metrics have a configuration that restricts model request payload to 100 KB. You can increase the payload size if required.Example - Model training and deployment (Iris)This topic uses Cloudera AI's built-in Python template project to walk you through an end-to-end example where we use experiments to develop and train a model, and then deploy it using Cloudera AI.
