AI for EverybodyLearn the intelligence era with better judgment.

Your progress: 0 complete

Glossary My Map Account

Glossary My Map AccountQuick answers Choose my route History of AI Sources About this platform

You are here

Current learning moveOrient

Follow a question, situation, concept, person, or historical thread.

Do nowRead, decide, then check

Your guide

Training builds behavior; inference uses it

Your learning move

OrientEnter through a question or situation.

Why this page exists

Explore is the connected atlas. It shows several ways to approach the same subject without pretending that every learner needs the same sequence.

What you will leave with

During training, a system adjusts model parameters using data, objectives, and feedback. During inference, the trained model receives a new input and produces an output.

Before you begin

No prerequisite is required for topic and Quick Answer entrances.

Do this now

Choose the entrance that matches what you already have: a question, little time, a situation, or a term.

Do now on this pageRead, decide, then checkThe plan states the outcome, time, prerequisite, immediate action, and prepared continuation.Another valid routeGet a Quick AnswerUse a short orientation before deciding whether to go deeper.

Account

Quick answer · 4 min

Training builds behavior; inference uses it

What is the difference between training and inference?

Plan for this page

Get a clear answer, then check that you can use it.

You will leave with: Training adjusts the model; inference applies it. Verify product retention and reuse separately.
Time: 4 min
Before you begin: No prerequisite. Start with your question.
Do this now: Read the direct answer before the example and active check.

Prepared next stepStudy models, training, and inference

During training, a system adjusts model parameters using data, objectives, and feedback. During inference, the trained model receives a new input and produces an output.

A conversation may add temporary context during inference without retraining the model. Product providers may later use feedback for separate training, depending on documented conditions.

Example

Learning and using a spam filter

Training fits patterns from labeled messages; inference classifies a newly arrived message.

Important limit

Product data flows are separate questions

The conceptual distinction does not tell you whether a provider stores prompts or later uses them for training; check the exact service.

Interactive model

Separate before deciding

Selected layerTraining

Changes model parameters through an optimization process.

Read the complete text alternative

TrainingChanges model parameters through an optimization process.
InferenceRuns current parameters on a new input.

Useful distinctions

Separate the questions before judging

Training

Changes model parameters through an optimization process.

Inference

Runs current parameters on a new input.

Active check

A chatbot answers your new question with its current model. What is happening?

Study models, training, and inference

Sources and scope

What supports this object

Established

Machine Learning GlossaryGoogle for Developers

Reference terminology; specific training and serving systems vary.

Open source

Loading page content...

…