Explanation: ADS supports hyperparameter tuning techniques like Grid Search and Random Search, which systematically or randomly explore the parameter space to optimize model performance. When using ADS AutoML, how are models compared for selection?
Explanation: AutoML is powerful for quick iteration, but it offers less control to advanced users who may want to manually choose algorithms, set parameters, or use custom pipelines for complex needs. Which components are required to save a model using ADSModel.from_estimator()? (Choose two)
Explanation: The Model Deployment service in OCI Data Science provisions compute resources and exposes endpoints for serving cataloged models. It manages autoscaling and deployment infrastructure. What must be done before deploying a model in OCI Data Science? (Choose two)
D. Converts models to GPU -compatible formats Explanation: Autoscaling allows deployed models to automatically adjust resource allocation based on demand. This optimizes cost and performance by scaling up under heavy load and down when idle. How does the ADS SDK support invoking a deployed model programmatically? (Choose two)
Detailed Answer in Step -by-Step Solution: Objective: Identify an open model format for cross -platform ML model execution. Evaluate Options: A . PySpark: A big data framework, not a model format. B . PyTorch: An ML framework with its own format, not inherently cross -platform without conversion. C . TensorFlow: An ML framework with its SavedModel format, not universally open across platforms. D . ONNX: Open Neural Network Exchange, an open -source format for model interoperability across frameworks. Reasoning: ONNX is designed for portability (e.g., convert PyTorch to ONNX, run in TensorFlow), unlike framework -specific options. Conclusion: D is the correct choice. ONNX (D) is “an open -source model format that enables interoperability between ML frameworks like PyTorch and TensorFlow,” per OCI documentation. PySpark (A) is a processing tool, while PyTorch (B) and TensorFlow (C) are frameworks with native formats —only ONNX ensures cross -platform compatibility. : Oracle Cloud Infrastructure Data Science Documentation, "Supported Model Formats". Where are OCI secrets stored?
What does OCI Generative AI provide as part of its integration with Data Science?
