agent-mlflow-skill-inference-test

To test model inference and schema compliance before deployment, you can run an inference test that validates the predict contract using an input_example.json file. This verifies the model's response schema and confirms the predict function works correctly.

Local model inference testing is the process of validating a model's predict function and response schema locally before deployment. It uses an input example to check the predict contract, ensuring the model returns outputs that adhere to the expected schema.

You validate your model's response schema by executing an inference test with a provided input_example.json. The test checks the predict function's functionality and ensures the returned response strictly adheres to the expected schema structure.

Yes, you need an input_example.json file to test your local model. The inference test uses this specific input file to check the functionality of the predict function and verify that the resulting output meets the required schema compliance.

Yes, this inference testing process includes entry point identification to determine the correct inference entrypoint for your model. It automatically locates and validates the predict contract to ensure the model is ready for deployment.

Schema errors during testing occur when the model's predict function returns a response that fails to adhere to the expected schema. Running an inference test with an input_example.json identifies these compliance issues and verifies the predict contract before deployment.