2.0 KiB
Model Training and Serving
"Model Training" refers to the phase in the Machine Learning (ML) pipeline where we teach a machine learning model how to make predictions by providing it with data. This process begins with feeding the model a training dataset, which it uses to learn and understand patterns or perform computations. The model's performance is then evaluated by comparing its prediction outputs with the actual results. Various algorithms can be used in the model training process. The choice of algorithm usually depends on the task, the data available, and the requirements of the project. It is worth noting that the model training stage can be computationally expensive particularly when dealing with large datasets or complex models.
Decisions depend on the organization's infrastructure.
- Repository Suggestion: ML Deployment k8s Fast API
- Tutorial Suggestions: ML deployment with k8s FastAPI, Building an ML app with FastAPI, Basic Kubeflow pipeline, Building and deploying ML pipelines, KServe tutorial