AI inference
Applying an already trained model to inputs to obtain a result.
What is AI inference?
AI inference applies a trained model to inputs. The result may be a prediction, class assignment or generated text. It uses learned model capabilities; a new input does not automatically mean new training. Depending on the design, inference can take place locally or on a server, individually or in batches.
In a fictional product workflow, a trained model proposes an item category for a new photo. That is inference; later adapting the model with appropriate training data would be a different operation. The team establishes how the proposal is checked and corrected when unclear. Response time, required resources and behaviour when no result arrives also belong in planning this processing step.
Sources
Related terms
- Machine learningModels learn patterns from data and apply them to new inputs.
- Neural networkA model of connected computational nodes with adjustable parameters.
- Fine-tuningFine-tuning adapts a pretrained model through further training. Supplying current documents at runtime is a different operation.
- LLM evaluationLLM evaluation tests model outputs against defined tasks and criteria. One good example does not establish overall quality.
- Model cardA model card documents an AI model’s purpose, foundations, evaluation conditions and known limitations for a particular version.
- AI readinessAI readiness concerns the foundations of an AI project: objectives, usable data, skills, responsibilities and an operating workflow.
Articles on this topic
AI & Automation with ASES
AI assistants and automated workflows