~/azure-gh-600 · Perform Evaluation, Error Analysis, and Tuning ▊
← GitHub Certified: Agentic AI Developer
AZURE · objective · 17% of the exam
Perform Evaluation, Error Analysis, and Tuning — GitHub Certified: Agentic AI Developer
The official AZURE documentation our Perform Evaluation, Error Analysis, and Tuning practice questions are cited to. Review the primary sources, then practise.
Official references for this objective
-
Microsoft — azure.ai.evaluation.MeteorScoreEvaluator class | Microsoft Learn
It addresses limitations of other metrics like BLEU by considering synonyms, stemming, and paraphrasing.
-
Microsoft — azure.ai.evaluation.ResponseCompletenessEvaluator class | Microsoft Learn
Evaluates the extent to which a given response contains all necessary and relevant information with respect to the provided ground truth.
-
Microsoft — azure.ai.evaluation.SelfHarmEvaluator class | Microsoft Learn
Evaluates self-harm score for a given query and response or a multi-turn conversation.
-
Quality.
-
List of output metrics produced by this evaluator.
-
The LLM model to use for rubric generation (e.g., 'gpt-4o'). Required — users must provide their own model rather than relying on service-owned capacity.
-
Microsoft — azure.ai.evaluation.RougeType enum | Microsoft Learn
Enumeration of ROUGE (Recall-Oriented Understudy for Gisting Evaluation) types.
-
Risk & Safety.
-
Prompt-based definition.