Build, Train, Deploy, and Query an MNIST Image Classifier on AWS
- Hyperparameter Optimization.
- Objective: Minimize Test Loss.
- Perfom hyperparameter search that satisfies this objective.
- Training.
- Train a custom CNN based model using the best hyperparameters from HPO step above.
- Deployment.
-
Method 1 (Basic Model Loader then Query in Notebook)
-
Method 2 (Extensive Process then Query with a Lambda Function). Tagged as _v2 in the code.
Built with ❤️ by martinoywa